SherbyNet - /page/wikibook/entry.mu

Micron Document

<!DOCTYPE html>

<title>ImageNet</title>

</head>

ImageNet

</h1>

</header>

The ImageNet project is a large visual <a href="Database" title="Database">database</a> designed for use in <a href="Outline_of_object_recognition" title="Outline of object recognition">visual object recognition software</a> research. More than 14 million<a href="#cite_note-New_Scientist-1">[1]</a><a href="#cite_note-nytimes_2012-2">[2]</a> images have been hand-annotated by the project to indicate what objects are pictured and in at least one million of the images, bounding boxes are also provided.<a href="#cite_note-3">[3]</a> ImageNet contains more than 20,000 categories,<a href="#cite_note-nytimes_2012-2">[2]</a> with a typical category, such as "balloon" or "strawberry", consisting of several hundred images.<a href="#cite_note-economist-4">[4]</a> The database of annotations of third-party image <a href="URL" title="URL">URLs</a> is freely available directly from ImageNet, though the actual images are not owned by ImageNet.<a href="#cite_note-5">[5]</a> Since 2010, the ImageNet project runs an annual software contest, the ImageNet Large Scale Visual Recognition Challenge (<a href="#History_of_the_ImageNet_challenge">ILSVRC</a>), where software programs compete to correctly classify and detect objects and scenes. The challenge uses a "trimmed" list of one thousand non-overlapping classes.<a href="#cite_note-ILJVRC-2015-6">[6]</a>

<div class="mw-heading mw-heading2"><h2 id="History">History</h2></div>

AI researcher <a href="Fei-Fei_Li" title="Fei-Fei Li">Fei-Fei Li</a> began working on the idea for ImageNet in 2006. At a time when most AI research focused on models and algorithms, Li wanted to expand and improve the data available to train AI algorithms.<a href="#cite_note-WiredQuest-7">[7]</a> In 2007, Li met with Princeton professor <a href="Christiane_Fellbaum" title="Christiane Fellbaum">Christiane Fellbaum</a>, one of the creators of <a href="WordNet" title="WordNet">WordNet</a>, to discuss the project. As a result of this meeting, Li went on to build ImageNet starting from the roughly 22,000 nouns of WordNet and using many of its features.<a href="#cite_note-Gershgorn-8">[8]</a> She was also inspired by a 1987 estimate<a href="#cite_note-9">[9]</a> that the average person recognizes roughly 30,000 different kinds of objects.<a href="#cite_note-:1-10">[10]</a>

As an assistant professor at <a href="Princeton_University" title="Princeton University">Princeton</a>, Li assembled a team of researchers to work on the ImageNet project. They used <a href="Amazon_Mechanical_Turk" title="Amazon Mechanical Turk">Amazon Mechanical Turk</a> to help with the classification of images. Labeling started in July 2008 and ended in April 2010. It took 49K workers from 167 countries filtering and labeling over 160M candidate images.<a href="#cite_note-11">[11]</a><a href="#cite_note-Gershgorn-8">[8]</a><a href="#cite_note-:7-12">[12]</a> They had enough budget to have each of the 14 million images labelled three times.<a href="#cite_note-:1-10">[10]</a>

The original plan called for 10,000 images per category, for 40,000 categories at 400 million images, each verified 3 times. They found that humans can classify at most 2 images/sec. At this rate, it was estimated to take 19 human-years of labor (without rest).<a href="#cite_note-:5-13">[13]</a>

They presented their database for the first time as a poster at the 2009 <a href="Conference_on_Computer_Vision_and_Pattern_Recognition" title="Conference on Computer Vision and Pattern Recognition">Conference on Computer Vision and Pattern Recognition</a> (CVPR) in Florida, titled "ImageNet: A Preview of a Large-scale Hierarchical Dataset".<a href="#cite_note-14">[14]</a><a href="#cite_note-Gershgorn-8">[8]</a><a href="#cite_note-:2-15">[15]</a><a href="#cite_note-16">[16]</a> The poster was reused at Vision Sciences Society 2009.<a href="#cite_note-17">[17]</a>

In 2009, Alex Berg suggested adding object localization as a task. Li approached <a rel="nofollow" class="external text" href="http://host.robots.ox.ac.uk/pascal/VOC/">PASCAL Visual Object Classes</a> contest in 2009 for a collaboration. It resulted in the subsequent <a href="ImageNet_Large_Scale_Visual_Recognition_Challenge" class="mw-redirect" title="ImageNet Large Scale Visual Recognition Challenge">ImageNet Large Scale Visual Recognition Challenge</a> starting in 2010, which has 1000 classes and object localization, as compared to <a rel="nofollow" class="external text" href="http://host.robots.ox.ac.uk/pascal/VOC/">PASCAL VOC</a> which had just 20 classes and 19,737 images (in 2010).<a href="#cite_note-ILJVRC-2015-6">[6]</a><a href="#cite_note-Gershgorn-8">[8]</a>

<div class="mw-heading mw-heading3"><h3 id="Significance_for_deep_learning">Significance for deep learning</h3></div>

On 30 September 2012, a <a href="Convolutional_neural_network" title="Convolutional neural network">convolutional neural network</a> (CNN) called <a href="AlexNet" title="AlexNet">AlexNet</a><a href="#cite_note-:0-18">[18]</a> achieved a top-5 error of 15.3% in the ImageNet 2012 Challenge, more than 10.8 percentage points lower than that of the runner-up. Using convolutional neural networks was feasible due to the use of <a href="Graphics_processing_unit" title="Graphics processing unit">graphics processing units</a> (GPUs) during training,<a href="#cite_note-:0-18">[18]</a> an essential ingredient of the <a href="Deep_learning" title="Deep learning">deep learning</a> revolution. According to <a href="The_Economist" title="The Economist">The Economist</a>, "Suddenly people started to pay attention, not just within the AI community but across the technology industry as a whole."<a href="#cite_note-economist-4">[4]</a><a href="#cite_note-19">[19]</a><a href="#cite_note-20">[20]</a>

In 2015, AlexNet was outperformed by <a href="Microsoft" title="Microsoft">Microsoft</a>'s <a href="ResNets" class="mw-redirect" title="ResNets">very deep CNN</a> with over 100 layers, which won the ImageNet 2015 contest, having 3.57% error on the test set.<a href="#cite_note-microsoft2015-21">[21]</a>

<a href="Andrej_Karpathy" title="Andrej Karpathy">Andrej Karpathy</a> estimated in 2014 that with concentrated effort, he could reach 5.1% Hit@5 error rate, and ~10 people from his lab reached ~12-13% with less effort.<a href="#cite_note-22">[22]</a><a href="#cite_note-23">[23]</a> It was estimated that with maximal effort, a human could reach 2.4%.<a href="#cite_note-ILJVRC-2015-6">[6]</a>

<div class="mw-heading mw-heading2"><h2 id="Dataset">Dataset</h2></div>

ImageNet <a href="Crowdsources" class="mw-redirect" title="Crowdsources">crowdsources</a> its annotation process. Image-level annotations indicate the presence or absence of an object class in an image, such as "there are tigers in this image" or "there are no tigers in this image". Object-level annotations provide a bounding box around the (visible part of the) indicated object. ImageNet uses a variant of the broad <a href="WordNet" title="WordNet">WordNet</a> schema to categorize objects, augmented with 120 categories of <a href="Dog_breeds" class="mw-redirect" title="Dog breeds">dog breeds</a> to showcase fine-grained classification.<a href="#cite_note-ILJVRC-2015-6">[6]</a>

In 2012, ImageNet was the world's largest academic user of <a href="Amazon_Mechanical_Turk" title="Amazon Mechanical Turk">Mechanical Turk</a>. The average worker identified 50 images per minute.<a href="#cite_note-nytimes_2012-2">[2]</a>

The original plan of the full ImageNet would have roughly 50M clean, diverse and full resolution images spread over approximately 50K synsets.<a href="#cite_note-:2-15">[15]</a> This was not achieved.

The summary statistics given on April 30, 2010:<a href="#cite_note-24">[24]</a>

<ul><li>Total number of non-empty synsets: 21841</li>

<li>Total number of images: 14,197,122</li>

<li>Number of images with bounding box annotations: 1,034,908</li>

<li>Number of synsets with SIFT features: 1000</li>

<li>Number of images with SIFT features: 1.2 million</li></ul>

<div class="mw-heading mw-heading3"><h3 id="Categories">Categories</h3></div>

The categories of ImageNet were filtered from the WordNet concepts. Each concept, since it can contain multiple synonyms (for example, "kitty" and "young cat"), so each concept is called a "synonym set" or "<a href="Synset" title="Synset">synset</a>". There were more than 100,000 synsets in WordNet 3.0, majority of them are nouns (80,000+). The ImageNet dataset filtered these to 21,841 synsets that are <a href="Count_noun" title="Count noun">countable nouns</a> that can be visually illustrated.

Each synset in WordNet 3.0 has a "WordNet ID" (wnid), which is a concatenation of <a href="Part_of_speech" title="Part of speech">part of speech</a> and an "offset" (a <a href="Unique_key" title="Unique key">unique identifying number</a>). Every wnid starts with "n" because ImageNet only includes <a href="Noun" title="Noun">nouns</a>. For example, the wnid of synset "<a href="Dog" title="Dog">dog, domestic dog, Canis familiaris</a>" is "n02084071".<a href="#cite_note-25">[25]</a>

The categories in ImageNet fall into 9 levels, from level 1 (such as "mammal") to level 9 (such as "German shepherd").<a href="#cite_note-:5-13">[13]</a>

<div class="mw-heading mw-heading3"><h3 id="Image_format">Image format</h3></div>

The images were scraped from online image search (<a href="Google_Search" title="Google Search">Google</a>, <a href="Picsearch" title="Picsearch">Picsearch</a>, <a href="MSN" title="MSN">MSN</a>, <a href="Yahoo" title="Yahoo">Yahoo</a>, <a href="Flickr" title="Flickr">Flickr</a>, etc) using synonyms in multiple languages. For example: German shepherd, German police dog, German shepherd dog, Alsatian, ovejero alemán, pastore tedesco, 德国牧羊犬.<a href="#cite_note-:4-26">[26]</a>

ImageNet consists of images in <a href="RGB_color_model" title="RGB color model">RGB</a> format with varying resolutions. For example, in ImageNet 2012, "fish" category, the resolution ranges from 4288 x 2848 to 75 x 56. In machine learning, these are typically preprocessed into a standard constant resolution, and whitened, before further processing by neural networks.

For example, in PyTorch, ImageNet images are by default normalized by dividing the pixel values so that they fall between 0 and 1, then subtracting by [0.485, 0.456, 0.406], then dividing by [0.229, 0.224, 0.225]. These are the mean and standard deviations for ImageNet, so this <a href="Whitening_transformation" title="Whitening transformation">whitens</a> the input data.<a href="#cite_note-27">[27]</a>

<div class="mw-heading mw-heading3"><h3 id="Labels_and_annotations">Labels and annotations</h3></div>

Each image is labelled with exactly one wnid.

Dense <a href="Scale-invariant_feature_transform" title="Scale-invariant feature transform">SIFT features</a> (raw SIFT descriptors, quantized codewords, and coordinates of each descriptor/codeword) for ImageNet-1K were available for download, designed for <a href="Bag_of_visual_words" class="mw-redirect" title="Bag of visual words">bag of visual words</a>.<a href="#cite_note-28">[28]</a>

The bounding boxes of objects were available for about 3000 popular synsets<a href="#cite_note-29">[29]</a> with on average 150 images in each synset.<a href="#cite_note-30">[30]</a>

Furthermore, some images have attributes. They released 25 attributes for ~400 popular synsets:<a href="#cite_note-31">[31]</a><a href="#cite_note-32">[32]</a>

<ul><li>Color: black, blue, brown, gray, green, orange, pink, red, violet, white, yellow</li>

<li>Pattern: spotted, striped</li>

<li>Shape: long, round, rectangular, square</li>

<li>Texture: furry, smooth, rough, shiny, metallic, vegetation, wooden, wet</li></ul>

<div class="mw-heading mw-heading3"><h3 id="ImageNet-21K">ImageNet-21K</h3></div>

The full original dataset is referred to as ImageNet-21K. ImageNet-21k contains 14,197,122 images divided into 21,841 classes. Some papers round this up and name it ImageNet-22k.<a href="#cite_note-:3-33">[33]</a>

The full ImageNet-21k was released in Fall of 2011, as <code>fall11_whole.tar</code>. There is no official train-validation-test split for ImageNet-21k. Some classes contain only 1-10 samples, while others contain thousands.<a href="#cite_note-:3-33">[33]</a>

<div class="mw-heading mw-heading3"><h3 id="ImageNet-1K">ImageNet-1K</h3></div>

There are various subsets of the ImageNet dataset used in various context, sometimes referred to as "versions".<a href="#cite_note-:0-18">[18]</a>

One of the most highly used subsets of ImageNet is the "ImageNet Large Scale Visual Recognition Challenge (ILSVRC) 2012–2017 image classification and localization dataset". This is also referred to in the research literature as ImageNet-1K or ILSVRC2017, reflecting the original ILSVRC challenge that involved 1,000 classes. ImageNet-1K contains 1,281,167 training images, 50,000 validation images and 100,000 test images.<a href="#cite_note-34">[34]</a>

Each category in ImageNet-1K is a leaf category, meaning that there are no child nodes below it, unlike ImageNet-21K. For example, in ImageNet-21K, there are some images categorized as simply "mammal", whereas in ImageNet-1K, there are only images categorized as things like "German shepherd", since there are no child-words below "German shepherd".<a href="#cite_note-:4-26">[26]</a>

<div class="mw-heading mw-heading3"><h3 id="Later_developments">Later developments</h3></div>

In the WordNet they built ImageNet on, there were 2832 synsets in the "person" subtree. During 2018--2020 period, they removed the download of the ImageNet-21k as they went through extensive filtering in these person synsets. Out of these 2832 synsets, 1593 were deemed "potentially offensive". Out of the remaining 1239, 1081 were deemed not really "visual". The result was that only 158 synsets remained. Of these, only 139 contained more than 100 images for "further exploration".<a href="#cite_note-:7-12">[12]</a><a href="#cite_note-:6-35">[35]</a><a href="#cite_note-36">[36]</a>

In 2021 winter, ImageNet-21k was updated. 2702 categories in the "person" subtree were removed to prevent "problematic behaviors" in a trained model. The result was that only 130 synsets in "person" subtree remained. Furthermore, in 2021, ImageNet-1k was updated by blurring out faces appearing in the 997 non-person categories. They found, out of all 1,431,093 images in ImageNet-1k, 243,198 images (17%) contain at least one face. And the total number of faces adds up to 562,626. They found training models on the dataset with these faces blurred caused minimal loss in performance.<a href="#cite_note-37">[37]</a><a href="#cite_note-38">[38]</a>

ImageNet-C is an adversarially perturbed version of ImageNet constructed in 2019.<a href="#cite_note-39">[39]</a>

ImageNetV2 was a new dataset containing three test sets with 10,000 each, constructed by the same methodology as the original ImageNet.<a href="#cite_note-40">[40]</a>

ImageNet-21K-P was a filtered and cleaned subset of ImageNet-21K, with 12,358,688 images from 11,221 categories. All Images were resized to 224 x 224px.<a href="#cite_note-:3-33">[33]</a>

<caption>Table of datasets

</caption>

<th>Name

</th>

<th>Published

</th>

<th>Classes

</th>

<th>Training

</th>

<th>Validation

</th>

<th>Test

</th>

<th>Size

</th></tr>

<tr>

<td>PASCAL VOC

</td>

<td>2005

</td>

<td>20

</td>

<td>

</td>

<td>

</td>

<td>

</td>

<td>

</td></tr>

<tr>

<td>ImageNet-1K

</td>

<td>2009

</td>

<td>1,000

</td>

<td>1,281,167

</td>

<td>50,000

</td>

<td>100,000

</td>

<td>130 GB

</td></tr>

<tr>

<td>ImageNet-21K

</td>

<td>2011

</td>

<td>21,841

</td>

<td>14,197,122

</td>

<td>

</td>

<td>

</td>

<td>1.31 TB

</td></tr>

<tr>

<td>ImageNetV2

</td>

<td>2019

</td>

<td>

</td>

<td>

</td>

<td>

</td>

<td>30,000

</td>

<td>

</td></tr>

<tr>

<td>ImageNet-21K-P

</td>

<td>2021

</td>

<td>11,221

</td>

<td>11,797,632

</td>

<td>

</td>

<td>561,052

</td>

</td></tr></tbody></table>

<div class="mw-heading mw-heading2"><h2 id="History_of_the_ImageNet_challenge">History of the ImageNet challenge</h2></div>

The ILSVRC aims to "follow in the footsteps" of the smaller-scale <a rel="nofollow" class="external text" href="http://host.robots.ox.ac.uk/pascal/VOC/">PASCAL VOC</a> challenge, established in 2005, which contained only about 20,000 images and twenty object classes.<a href="#cite_note-ILJVRC-2015-6">[6]</a> To "democratize" ImageNet, Fei-Fei Li proposed to the <a rel="nofollow" class="external text" href="http://host.robots.ox.ac.uk/pascal/VOC/">PASCAL VOC</a> team a collaboration, beginning in 2010, where research teams would evaluate their algorithms on the given data set, and compete to achieve higher accuracy on several visual recognition tasks.<a href="#cite_note-Gershgorn-8">[8]</a>

The resulting annual competition is now known as the ImageNet Large Scale Visual Recognition Challenge (ILSVRC). The ILSVRC uses a "trimmed" list of only 1000 image categories or "classes", including 90 of the 120 dog breeds classified by the full ImageNet schema.<a href="#cite_note-ILJVRC-2015-6">[6]</a>

The 2010s saw dramatic progress in image processing.

The first competition in 2010 had 11 participating teams. The winning team was a linear <a href="Support_vector_machine" title="Support vector machine">support vector machine</a> (SVM). The features are a dense grid of <a href="Histogram_of_oriented_gradients" title="Histogram of oriented gradients">HoG</a> and <a href="Local_binary_patterns" title="Local binary patterns">LBP</a>, sparsified by local coordinate coding and pooling.<a href="#cite_note-41">[41]</a> It achieved 52.9% in classification accuracy and 71.8% in top-5 accuracy. It was trained for 4 days on three 8-core machines (dual quad-core 2 GHz <a href="Xeon" title="Xeon">Intel Xeon</a> CPU).<a href="#cite_note-42">[42]</a>

The second competition in 2011 had fewer teams, with another SVM winning at top-5 error rate 25%.<a href="#cite_note-:1-10">[10]</a> The winning team was XRCE by Florent Perronnin, Jorge Sanchez. The system was another linear SVM, running on quantized<a href="#cite_note-43">[43]</a> <a href="Fisher_kernel" title="Fisher kernel">Fisher vectors</a>.<a href="#cite_note-44">[44]</a><a href="#cite_note-45">[45]</a> It achieved 74.2% in top-5 accuracy.

In 2012, a deep <a href="Convolutional_neural_network" title="Convolutional neural network">convolutional neural net</a> called <a href="AlexNet" title="AlexNet">AlexNet</a> achieved 84.7% in top-5 accuracy, a great leap forward.<a href="#cite_note-46">[46]</a> The second place was by Oxford VGG, which uses the previous generic architecture of SVM, SIFT, color statistics, Fisher vectors, etc.<a href="#cite_note-47">[47]</a> In the next couple of years, top-5 accuracy grew to above 90%. While the 2012 breakthrough "combined pieces that were all there before", the dramatic quantitative improvement marked the start of an industry-wide artificial intelligence boom.<a href="#cite_note-economist-4">[4]</a>

In 2013, most high-ranking entries used convolutional neural networks. The winning entry for object localization was the <a href="You_Only_Look_Once#OverFeat" title="You Only Look Once">OverFeat</a>, an architecture for simultaneous object classification and localization.<a href="#cite_note-48">[48]</a> The winning entry for classification was an ensemble of multiple CNNs by Clarifai.<a href="#cite_note-ILJVRC-2015-6">[6]</a>

By 2014, more than 50 institutions participated in the ILSVRC.<a href="#cite_note-ILJVRC-2015-6">[6]</a> The winning entry for classification was <a href="Inception_(deep_learning_architecture)" title="Inception (deep learning architecture)">GoogLeNet</a>.<a href="#cite_note-szegedy-49">[49]</a> The winning entry for localization was <a href="VGGNet" title="VGGNet">VGGNet</a>. In 2017, 29 of 38 competing teams had greater than 95% accuracy.<a href="#cite_note-50">[50]</a> In 2017 ImageNet stated it would roll out a new, much more difficult challenge in 2018 that involves classifying 3D objects using natural language. Because creating 3D data is more costly than annotating a pre-existing 2D image, the dataset is expected to be smaller. The applications of progress in this area would range from robotic navigation to <a href="Augmented_reality" title="Augmented reality">augmented reality</a>.<a href="#cite_note-New_Scientist-1">[1]</a>

In 2015, the winning entry was <a href="Residual_neural_network" title="Residual neural network">ResNet</a>, which exceeded human performance.<a href="#cite_note-microsoft2015-21">[21]</a><a href="#cite_note-51">[51]</a> However, as one of the challenge's organizers, <a href="Olga_Russakovsky" title="Olga Russakovsky">Olga Russakovsky</a>, pointed out in 2015, the ILSVRC is over only 1000 categories; humans can recognize a larger number of categories, and also (unlike the programs) can judge the context of an image.<a href="#cite_note-52">[52]</a>

In 2016, the winning entry was CUImage, an <a href="Ensemble_learning" title="Ensemble learning">ensemble</a> model of 6 networks: Inception v3, Inception v4, Inception ResNet v2, ResNet 200, Wide ResNet 68, and Wide ResNet 3.<a href="#cite_note-53">[53]</a> The runner-up was ResNeXt, which combines the Inception module with ResNet.<a href="#cite_note-54">[54]</a>

In 2017, the winning entry was the <a href="Squeeze-and-Excitation_Network" class="mw-redirect" title="Squeeze-and-Excitation Network">Squeeze-and-Excitation Network</a> (SENet), reducing the top-5 error to 2.251%.<a href="#cite_note-55">[55]</a>

The organizers of the competition stated in 2017 that the 2017 competition would be the last one, since the benchmark has been solved and no longer posed a challenge. They also stated that they would organize a new competition on 3D images.<a href="#cite_note-New_Scientist-1">[1]</a> However, such a competition never materialized.

<div class="mw-heading mw-heading2"><h2 id="Bias_in_ImageNet">Bias in ImageNet</h2></div>

It is estimated that over 6% of labels in the ImageNet-1k validation set are wrong.<a href="#cite_note-56">[56]</a> It is also found that around 10% of ImageNet-1k contains ambiguous or erroneous labels, and that, when presented with a model's prediction and the original ImageNet label, human annotators prefer the prediction of a state of the art model in 2020 trained on the original ImageNet, suggesting that ImageNet-1k has been saturated.<a href="#cite_note-57">[57]</a>

A study of the history of the multiple layers (<a href="Taxonomy_(general)" class="mw-redirect" title="Taxonomy (general)">taxonomy</a>, object classes and labeling) of ImageNet and WordNet in 2019 described how <a href="Algorithmic_bias" title="Algorithmic bias">bias</a> is deeply embedded in most classification approaches for all sorts of images.<a href="#cite_note-58">[58]</a><a href="#cite_note-59">[59]</a><a href="#cite_note-60">[60]</a><a href="#cite_note-61">[61]</a> ImageNet is working to address various sources of bias.<a href="#cite_note-62">[62]</a>

One downside of WordNet use is the categories may be more "elevated" than would be optimal for ImageNet: "Most people are more interested in Lady Gaga or the iPod Mini than in this rare kind of <a href="Diplodocus" title="Diplodocus">diplodocus</a>."

<ul><li><a href="Computer_vision" title="Computer vision">Computer vision</a></li>

<li><a href="List_of_datasets_for_machine_learning_research" class="mw-redirect" title="List of datasets for machine learning research">List of datasets for machine learning research</a></li>

<li><a href="WordNet" title="WordNet">WordNet</a></li></ul>

<div class="mw-heading mw-heading2"><h2 id="References">References</h2></div>

/* start https://en.wikipedia.org/ */

.mw-parser-output .reflist{margin-bottom:0.5em;list-style-type:decimal}@media screen{.mw-parser-output .reflist{font-size:90%}}.mw-parser-output .reflist .references{font-size:100%;margin-bottom:0;list-style-type:inherit}.mw-parser-output .reflist-columns-2{column-width:30em}.mw-parser-output .reflist-columns-3{column-width:25em}.mw-parser-output .reflist-columns{margin-top:0.3em}.mw-parser-output .reflist-columns ol{margin-top:0}.mw-parser-output .reflist-columns li{page-break-inside:avoid;break-inside:avoid-column}.mw-parser-output .reflist-upper-alpha{list-style-type:upper-alpha}.mw-parser-output .reflist-upper-roman{list-style-type:upper-roman}.mw-parser-output .reflist-lower-alpha{list-style-type:lower-alpha}.mw-parser-output .reflist-lower-greek{list-style-type:lower-greek}.mw-parser-output .reflist-lower-roman{list-style-type:lower-roman}

/* end https://en.wikipedia.org/ */

</style><div class="reflist reflist-columns references-column-width" style="column-width: 30em;">

/* start https://en.wikipedia.org/ */

.mw-parser-output cite.citation{font-style:inherit;word-wrap:break-word}.mw-parser-output .citation q{quotes:"\"""\"""'""'"}.mw-parser-output .citation:target{background-color:rgba(0,127,255,0.133)}.mw-parser-output .id-lock-free.id-lock-free a{background:url("./mw/Lock-green.svg")right 0.1em center/9px no-repeat}.mw-parser-output .id-lock-limited.id-lock-limited a,.mw-parser-output .id-lock-registration.id-lock-registration a{background:url("./mw/Lock-gray-alt-2.svg")right 0.1em center/9px no-repeat}.mw-parser-output .id-lock-subscription.id-lock-subscription a{background:url("./mw/Lock-red-alt-2.svg")right 0.1em center/9px no-repeat}.mw-parser-output .cs1-ws-icon a{background:url("./mw/Wikisource-logo.svg")right 0.1em center/12px no-repeat}body:not(.skin-timeless):not(.skin-minerva) .mw-parser-output .id-lock-free a,body:not(.skin-timeless):not(.skin-minerva) .mw-parser-output .id-lock-limited a,body:not(.skin-timeless):not(.skin-minerva) .mw-parser-output .id-lock-registration a,body:not(.skin-timeless):not(.skin-minerva) .mw-parser-output .id-lock-subscription a,body:not(.skin-timeless):not(.skin-minerva) .mw-parser-output .cs1-ws-icon a{background-size:contain;padding:0 1em 0 0}.mw-parser-output .cs1-code{color:inherit;background:inherit;border:none;padding:inherit}.mw-parser-output .cs1-hidden-error{display:none;color:var(--color-error,#d33)}.mw-parser-output .cs1-visible-error{color:var(--color-error,#d33)}.mw-parser-output .cs1-maint{display:none;color:#085;margin-left:0.3em}.mw-parser-output .cs1-kern-left{padding-left:0.2em}.mw-parser-output .cs1-kern-right{padding-right:0.2em}.mw-parser-output .citation .mw-selflink{font-weight:inherit}@media screen{.mw-parser-output .cs1-format{font-size:95%}html.skin-theme-clientpref-night .mw-parser-output .cs1-maint{color:#18911f}}@media screen and (prefers-color-scheme:dark){html.skin-theme-clientpref-os .mw-parser-output .cs1-maint{color:#18911f}}

/* end https://en.wikipedia.org/ */

</style><cite class="citation news cs1"><a rel="nofollow" class="external text" href="https://www.newscientist.com/article/2127131-new-computer-vision-challenge-wants-to-teach-robots-to-see-in-3d/">"New computer vision challenge wants to teach robots to see in 3D"</a>. New Scientist. 7 April 2017. Retrieved 3 February 2018.</cite>

</li>

<li id="cite_note-nytimes_2012-2">^ <a href="#cite_ref-nytimes_2012_2-0">a</a> <a href="#cite_ref-nytimes_2012_2-1">b</a> <a href="#cite_ref-nytimes_2012_2-2">c</a> <cite id="CITEREFMarkoff2012" class="citation news cs1">Markoff, John (19 November 2012). <a rel="nofollow" class="external text" href="https://www.nytimes.com/2012/11/20/science/for-web-images-creating-new-technology-to-seek-and-find.html">"For Web Images, Creating New Technology to Seek and Find"</a>. The New York Times. Retrieved 3 February 2018.</cite>

</li>

<li id="cite_note-3"><a href="#cite_ref-3">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://web.archive.org/web/20200907212153/http://image-net.org/about-stats.php">"ImageNet"</a>. 7 September 2020. Archived from <a rel="nofollow" class="external text" href="http://image-net.org/about-stats.php">the original</a> on 7 September 2020. Retrieved 11 October 2022.</cite>

</li>

<li id="cite_note-economist-4">^ <a href="#cite_ref-economist_4-0">a</a> <a href="#cite_ref-economist_4-1">b</a> <a href="#cite_ref-economist_4-2">c</a> <cite class="citation news cs1"><a rel="nofollow" class="external text" href="https://www.economist.com/news/special-report/21700756-artificial-intelligence-boom-based-old-idea-modern-twist-not">"From not working to neural networking"</a>. The Economist. 25 June 2016. Retrieved 3 February 2018.</cite>

</li>

<li id="cite_note-5"><a href="#cite_ref-5">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://image-net.org/about.php">"ImageNet Overview"</a>. ImageNet. Retrieved 15 October 2022.</cite>

</li>

<li id="cite_note-ILJVRC-2015-6">^ <a href="#cite_ref-ILJVRC-2015_6-0">a</a> <a href="#cite_ref-ILJVRC-2015_6-1">b</a> <a href="#cite_ref-ILJVRC-2015_6-2">c</a> <a href="#cite_ref-ILJVRC-2015_6-3">d</a> <a href="#cite_ref-ILJVRC-2015_6-4">e</a> <a href="#cite_ref-ILJVRC-2015_6-5">f</a> <a href="#cite_ref-ILJVRC-2015_6-6">g</a> <a href="#cite_ref-ILJVRC-2015_6-7">h</a> <cite id="CITEREFRussakovskyDengSuKrause2015" class="citation journal cs1">Russakovsky, Olga; Deng, Jia; Su, Hao; Krause, Jonathan; Satheesh, Sanjeev; Ma, Sean; Huang, Zhiheng; Karpathy, Andrej; Khosla, Aditya; Bernstein, Michael; Berg, Alexander C.; Fei-Fei, Li (1 December 2015). <a rel="nofollow" class="external text" href="https://doi.org/10.1007/s11263-015-0816-y">"ImageNet Large Scale Visual Recognition Challenge"</a>. International Journal of Computer Vision. 115 (3): 211–252. <a href="ArXiv_(identifier)" class="mw-redirect" title="ArXiv (identifier)">arXiv</a>:<a rel="nofollow" class="external text" href="https://arxiv.org/abs/1409.0575">1409.0575</a>. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1007%2Fs11263-015-0816-y">10.1007/s11263-015-0816-y</a>. <a href="ISSN_(identifier)" class="mw-redirect" title="ISSN (identifier)">ISSN</a> <a rel="nofollow" class="external text" href="https://search.worldcat.org/issn/1573-1405">1573-1405</a>.</cite>

</li>

<li id="cite_note-WiredQuest-7"><a href="#cite_ref-WiredQuest_7-0">^</a> <cite id="CITEREFHempel2018" class="citation magazine cs1">Hempel, Jesse (13 November 2018). <a rel="nofollow" class="external text" href="https://www.wired.com/story/fei-fei-li-artificial-intelligence-humanity/">"Fei-Fei Li's Quest to Make AI Better for Humanity"</a>. Wired. Retrieved 5 May 2019. <q>When Li, who had moved back to Princeton to take a job as an assistant professor in 2007, talked up her idea for ImageNet, she had a hard time getting faculty members to help out. Finally, a professor who specialized in computer architecture agreed to join her as a collaborator.</q></cite>

</li>

<li id="cite_note-Gershgorn-8">^ <a href="#cite_ref-Gershgorn_8-0">a</a> <a href="#cite_ref-Gershgorn_8-1">b</a> <a href="#cite_ref-Gershgorn_8-2">c</a> <a href="#cite_ref-Gershgorn_8-3">d</a> <a href="#cite_ref-Gershgorn_8-4">e</a> <cite id="CITEREFGershgorn2017" class="citation web cs1">Gershgorn, Dave (26 July 2017). <a rel="nofollow" class="external text" href="https://qz.com/1034972/the-data-that-changed-the-direction-of-ai-research-and-possibly-the-world/">"The data that transformed AI research—and possibly the world"</a>. Quartz. Atlantic Media Co. Retrieved 26 July 2017. <q>Having read about WordNet's approach, Li met with professor Christiane Fellbaum, a researcher influential in the continued work on WordNet, during a 2006 visit to Princeton.</q></cite>

</li>

<li id="cite_note-9"><a href="#cite_ref-9">^</a> <cite id="CITEREFBiederman1987" class="citation journal cs1"><a href="Irving_Biederman" title="Irving Biederman">Biederman, Irving</a> (1987). <a rel="nofollow" class="external text" href="https://dx.doi.org/10.1037//0033-295x.94.2.115">"Recognition-by-components: A theory of human image understanding"</a>. Psychological Review. 94 (2): 115–117. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1037%2F0033-295x.94.2.115">10.1037/0033-295x.94.2.115</a>. <a href="ISSN_(identifier)" class="mw-redirect" title="ISSN (identifier)">ISSN</a> <a rel="nofollow" class="external text" href="https://search.worldcat.org/issn/0033-295X">0033-295X</a>. <a href="PMID_(identifier)" class="mw-redirect" title="PMID (identifier)">PMID</a> <a rel="nofollow" class="external text" href="https://pubmed.ncbi.nlm.nih.gov/3575582">3575582</a>.</cite>

</li>

<li id="cite_note-:1-10">^ <a href="#cite_ref-:1_10-0">a</a> <a href="#cite_ref-:1_10-1">b</a> <a href="#cite_ref-:1_10-2">c</a> <cite id="CITEREFLee2024" class="citation web cs1">Lee, Timothy B. (11 November 2024). <a rel="nofollow" class="external text" href="https://arstechnica.com/ai/2024/11/how-a-stubborn-computer-scientist-accidentally-launched-the-deep-learning-boom/">"How a stubborn computer scientist accidentally launched the deep learning boom"</a>. Ars Technica. Retrieved 12 November 2024.</cite>

</li>

<li id="cite_note-11"><a href="#cite_ref-11">^</a> <cite id="CITEREFLiDeng2017" class="citation conference cs1">Li, Fei-Fei; Deng, Jia (2017). <a rel="nofollow" class="external text" href="https://image-net.org/static_files/files/imagenet_ilsvrc2017_v1.0.pdf">Where have we been? Where are we going?</a> (PDF). Beyond ImageNet Large Scale Visual Recognition Challenge, Workshop at CVPR 2017 (Presentation).</cite>

</li>

<li id="cite_note-:7-12">^ <a href="#cite_ref-:7_12-0">a</a> <a href="#cite_ref-:7_12-1">b</a> <cite id="CITEREFYangQinamiFei-FeiDeng2019" class="citation web cs1">Yang, Kaiyu; Qinami, Klint; Fei-Fei, Li; Deng, Jia; Russakovsky, Olga (17 September 2019). <a rel="nofollow" class="external text" href="https://www.image-net.org/update-sep-17-2019.php">"Towards Fairer Datasets: Filtering and Balancing the Distribution of the People Subtree in the ImageNet Hierarchy"</a>. image-net.org.</cite>

</li>

<li id="cite_note-:5-13">^ <a href="#cite_ref-:5_13-0">a</a> <a href="#cite_ref-:5_13-1">b</a> Li, F-F. ImageNet. "<a rel="nofollow" class="external text" href="https://web.archive.org/web/20130115112543/http://www.image-net.org/papers/ImageNet_2010.pdf">Crowdsourcing, benchmarking & other cool things</a>." CMU VASC Semin 16 (2010): 18-25.

</li>

<li id="cite_note-14"><a href="#cite_ref-14">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="http://tab.computer.org/pamitc/archive/cvpr2009/posters.html">"CVPR 2009: IEEE Computer Society Conference on Computer Vision and Pattern Recognition"</a>. tab.computer.org. Retrieved 13 November 2024.</cite>

</li>

<li id="cite_note-:2-15">^ <a href="#cite_ref-:2_15-0">a</a> <a href="#cite_ref-:2_15-1">b</a> <cite id="CITEREFDengDongSocherLi2009" class="citation cs2">Deng, Jia; Dong, Wei; Socher, Richard; Li, Li-Jia; Li, Kai; Fei-Fei, Li (2009), <a rel="nofollow" class="external text" href="https://web.archive.org/web/20210115185228/http://www.image-net.org/papers/imagenet_cvpr09.pdf">"ImageNet: A Large-Scale Hierarchical Image Database"</a> (PDF), 2009 conference on Computer Vision and Pattern Recognition, archived from <a rel="nofollow" class="external text" href="http://www.image-net.org/papers/imagenet_cvpr09.pdf">the original</a> (PDF) on 15 January 2021, retrieved 26 July 2017</cite>

</li>

<li id="cite_note-16"><a href="#cite_ref-16">^</a> <cite id="CITEREFLi2015" class="citation cs2">Li, Fei-Fei (23 March 2015), <a rel="nofollow" class="external text" href="https://www.ted.com/talks/fei_fei_li_how_we_re_teaching_computers_to_understand_pictures?language=en">How we're teaching computers to understand pictures</a>, retrieved 16 December 2018</cite>

</li>

<li id="cite_note-17"><a href="#cite_ref-17">^</a> Deng, Jia, et al. "<a rel="nofollow" class="external text" href="https://web.archive.org/web/20130115112451/http://www.image-net.org/papers/ImageNet_VSS2009.pdf">Construction and analysis of a large scale image ontology</a>." Vision Sciences Society 186.2 (2009).

</li>

<li id="cite_note-:0-18">^ <a href="#cite_ref-:0_18-0">a</a> <a href="#cite_ref-:0_18-1">b</a> <a href="#cite_ref-:0_18-2">c</a> <cite id="CITEREFKrizhevskySutskeverHinton2017" class="citation journal cs1">Krizhevsky, Alex; Sutskever, Ilya; Hinton, Geoffrey E. (June 2017). <a rel="nofollow" class="external text" href="https://papers.nips.cc/paper/4824-imagenet-classification-with-deep-convolutional-neural-networks.pdf">"ImageNet classification with deep convolutional neural networks"</a> (PDF). Communications of the ACM. 60 (6): 84–90. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1145%2F3065386">10.1145/3065386</a>. <a href="ISSN_(identifier)" class="mw-redirect" title="ISSN (identifier)">ISSN</a> <a rel="nofollow" class="external text" href="https://search.worldcat.org/issn/0001-0782">0001-0782</a>. <a href="S2CID_(identifier)" class="mw-redirect" title="S2CID (identifier)">S2CID</a> <a rel="nofollow" class="external text" href="https://api.semanticscholar.org/CorpusID:195908774">195908774</a>. Retrieved 24 May 2017.</cite>

</li>

<li id="cite_note-19"><a href="#cite_ref-19">^</a> <cite class="citation news cs1"><a rel="nofollow" class="external text" href="https://www.ft.com/content/4cc048f6-d5f4-11e7-a303-9060cb1e5f44">"Machines 'beat humans' for a growing number of tasks"</a>. Financial Times. 30 November 2017. Retrieved 3 February 2018.</cite>

</li>

<li id="cite_note-20"><a href="#cite_ref-20">^</a> <cite id="CITEREFGershgorn2018" class="citation web cs1">Gershgorn, Dave (18 June 2018). <a rel="nofollow" class="external text" href="https://qz.com/1307091/the-inside-story-of-how-ai-got-good-enough-to-dominate-silicon-valley/">"The inside story of how AI got good enough to dominate Silicon Valley"</a>. Quartz. Retrieved 10 December 2018.</cite>

</li>

<li id="cite_note-microsoft2015-21">^ <a href="#cite_ref-microsoft2015_21-0">a</a> <a href="#cite_ref-microsoft2015_21-1">b</a> <cite id="CITEREFHeZhangRenSun2016" class="citation book cs1">He, Kaiming; Zhang, Xiangyu; Ren, Shaoqing; Sun, Jian (2016). "Deep Residual Learning for Image Recognition". 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 770–778. <a href="ArXiv_(identifier)" class="mw-redirect" title="ArXiv (identifier)">arXiv</a>:<a rel="nofollow" class="external text" href="https://arxiv.org/abs/1512.03385">1512.03385</a>. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1109%2FCVPR.2016.90">10.1109/CVPR.2016.90</a>. <a href="ISBN_(identifier)" class="mw-redirect" title="ISBN (identifier)">ISBN</a> <bdi>978-1-4673-8851-1</bdi>. <a href="S2CID_(identifier)" class="mw-redirect" title="S2CID (identifier)">S2CID</a> <a rel="nofollow" class="external text" href="https://api.semanticscholar.org/CorpusID:206594692">206594692</a>.</cite>

</li>

<li id="cite_note-22"><a href="#cite_ref-22">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://web.archive.org/web/20150522055300/https://plus.google.com/+PierreSermanet/posts/6wZYMuXo8PU">"New community features for Google Chat and an update on Currents"</a>. Archived from <a rel="nofollow" class="external text" href="https://plus.google.com/+PierreSermanet/posts/6wZYMuXo8PU">the original</a> on 22 May 2015.</cite>

</li>

<li id="cite_note-23"><a href="#cite_ref-23">^</a> <cite id="CITEREFKarpathy2014" class="citation web cs1">Karpathy, Andrej (2 September 2014). <a rel="nofollow" class="external text" href="https://karpathy.github.io/2014/09/02/what-i-learned-from-competing-against-a-convnet-on-imagenet/">"What I learned from competing against a ConvNet on ImageNet"</a>. Andrej Karpathy blog.</cite>

</li>

<li id="cite_note-24"><a href="#cite_ref-24">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://web.archive.org/web/20130115112755/http://www.image-net.org/about-stats">"ImageNet Summary and Statistics (updated on April 30, 2010)"</a>. 15 January 2013. Archived from <a rel="nofollow" class="external text" href="http://www.image-net.org/about-stats">the original</a> on 15 January 2013. Retrieved 13 November 2024.</cite>

</li>

<li id="cite_note-25"><a href="#cite_ref-25">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://web.archive.org/web/20130122145752/http://www.image-net.org/download-API">"ImageNet API documentation"</a>. 22 January 2013. Archived from <a rel="nofollow" class="external text" href="http://www.image-net.org/download-API">the original</a> on 22 January 2013. Retrieved 13 November 2024.</cite>

</li>

<li id="cite_note-:4-26">^ <a href="#cite_ref-:4_26-0">a</a> <a href="#cite_ref-:4_26-1">b</a> Berg, Alex, Jia Deng, and L. Fei-Fei. "<a rel="nofollow" class="external text" href="https://www.image-net.org/static_files/files/pascal_ilsvrc.pdf">Large scale visual recognition challenge 2010</a>." November 2010.

</li>

<li id="cite_note-27"><a href="#cite_ref-27">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://github.com/openai/CLIP/issues/20">"std and mean for image normalization different from ImageNet · Issue #20 · openai/CLIP"</a>. GitHub. Retrieved 19 September 2024.</cite>

</li>

<li id="cite_note-28"><a href="#cite_ref-28">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://web.archive.org/web/20130405035300/http://www.image-net.org/download-features.php">"ImageNet"</a>. 5 April 2013. Archived from <a rel="nofollow" class="external text" href="http://www.image-net.org/download-features.php">the original</a> on 5 April 2013. Retrieved 13 November 2024.</cite>

</li>

<li id="cite_note-29"><a href="#cite_ref-29">^</a> <a rel="nofollow" class="external free" href="https://web.archive.org/web/20181030191122/http://www.image-net.org/api/text/imagenet.sbow.obtain_synset_list">https://web.archive.org/web/20181030191122/http://www.image-net.org/api/text/imagenet.sbow.obtain_synset_list</a>

</li>

<li id="cite_note-30"><a href="#cite_ref-30">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://web.archive.org/web/20130405005059/http://www.image-net.org/download-bboxes">"ImageNet"</a>. Archived from <a rel="nofollow" class="external text" href="http://www.image-net.org/download-bboxes">the original</a> on 5 April 2013.</cite>

</li>

<li id="cite_note-31"><a href="#cite_ref-31">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://web.archive.org/web/20191222152337/http://www.image-net.org/download-attributes">"ImageNet"</a>. Archived from <a rel="nofollow" class="external text" href="http://www.image-net.org/download-attributes">the original</a> on 22 December 2019.</cite>

</li>

<li id="cite_note-32"><a href="#cite_ref-32">^</a> <cite id="CITEREFRussakovskyFei-Fei2012" class="citation book cs1">Russakovsky, Olga; Fei-Fei, Li (2012). <a rel="nofollow" class="external text" href="https://link.springer.com/chapter/10.1007/978-3-642-35749-7_1">"Attribute Learning in Large-Scale Datasets"</a>. In Kutulakos, Kiriakos N. (ed.). Trends and Topics in Computer Vision. Lecture Notes in Computer Science. Vol. 6553. Berlin, Heidelberg: Springer. pp. 1–14. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1007%2F978-3-642-35749-7_1">10.1007/978-3-642-35749-7_1</a>. <a href="ISBN_(identifier)" class="mw-redirect" title="ISBN (identifier)">ISBN</a> <bdi>978-3-642-35749-7</bdi>.</cite>

</li>

<li id="cite_note-:3-33">^ <a href="#cite_ref-:3_33-0">a</a> <a href="#cite_ref-:3_33-1">b</a> <a href="#cite_ref-:3_33-2">c</a> <a href="#cite_ref-:3_33-3">d</a> <cite id="CITEREFRidnikBen-BaruchNoyZelnik-Manor2021" class="citation arxiv cs1">Ridnik, Tal; Ben-Baruch, Emanuel; Noy, Asaf; Zelnik-Manor, Lihi (5 August 2021). "ImageNet-21K Pretraining for the Masses". <a href="ArXiv_(identifier)" class="mw-redirect" title="ArXiv (identifier)">arXiv</a>:<a rel="nofollow" class="external text" href="https://arxiv.org/abs/2104.10972">2104.10972</a> [<a rel="nofollow" class="external text" href="https://arxiv.org/archive/cs.CV">cs.CV</a>].</cite>

</li>

<li id="cite_note-34"><a href="#cite_ref-34">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://www.image-net.org/download.php">"ImageNet"</a>. www.image-net.org. Retrieved 19 October 2022.</cite>

</li>

<li id="cite_note-:6-35"><a href="#cite_ref-:6_35-0">^</a> <cite id="CITEREFYangQinamiFei-FeiDeng2020" class="citation book cs1">Yang, Kaiyu; Qinami, Klint; Fei-Fei, Li; Deng, Jia; Russakovsky, Olga (27 January 2020). <a rel="nofollow" class="external text" href="https://dl.acm.org/doi/10.1145/3351095.3375709">"Towards fairer datasets: Filtering and balancing the distribution of the people subtree in the ImageNet hierarchy"</a>. Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency. ACM. pp. 547–558. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1145%2F3351095.3375709">10.1145/3351095.3375709</a>. <a href="ISBN_(identifier)" class="mw-redirect" title="ISBN (identifier)">ISBN</a> <bdi>978-1-4503-6936-7</bdi>.</cite>

</li>

<li id="cite_note-36"><a href="#cite_ref-36">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://www.nsf.gov/awardsearch/showAward?AWD_ID=1763642&HistoricalAwards=false">"NSF Award Search: Award # 1763642"</a>. www.nsf.gov. Retrieved 7 June 2025.</cite>

</li>

<li id="cite_note-37"><a href="#cite_ref-37">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://www.image-net.org/update-mar-11-2021.php">"An Update to the ImageNet Website and Dataset"</a>. www.image-net.org. Retrieved 13 November 2024.</cite>

</li>

<li id="cite_note-38"><a href="#cite_ref-38">^</a> <cite id="CITEREFYangYauFei-FeiDeng2022" class="citation journal cs1">Yang, Kaiyu; Yau, Jacqueline H.; Fei-Fei, Li; Deng, Jia; Russakovsky, Olga (28 June 2022). <a rel="nofollow" class="external text" href="https://proceedings.mlr.press/v162/yang22q.html">"A Study of Face Obfuscation in ImageNet"</a>. Proceedings of the 39th International Conference on Machine Learning. PMLR: 25313–25330.</cite>

</li>

<li id="cite_note-39"><a href="#cite_ref-39">^</a> <cite id="CITEREFHendrycksDietterich2019" class="citation arxiv cs1">Hendrycks, Dan; Dietterich, Thomas (2019). "Benchmarking Neural Network Robustness to Common Corruptions and Perturbations". <a href="ArXiv_(identifier)" class="mw-redirect" title="ArXiv (identifier)">arXiv</a>:<a rel="nofollow" class="external text" href="https://arxiv.org/abs/1903.12261">1903.12261</a> [<a rel="nofollow" class="external text" href="https://arxiv.org/archive/cs.LG">cs.LG</a>].</cite>

</li>

<li id="cite_note-40"><a href="#cite_ref-40">^</a> <cite id="CITEREFRechtRoelofsSchmidtShankar2019" class="citation journal cs1">Recht, Benjamin; Roelofs, Rebecca; Schmidt, Ludwig; Shankar, Vaishaal (24 May 2019). <a rel="nofollow" class="external text" href="https://proceedings.mlr.press/v97/recht19a.html">"Do ImageNet Classifiers Generalize to ImageNet?"</a>. Proceedings of the 36th International Conference on Machine Learning. PMLR: 5389–5400.</cite>

</li>

<li id="cite_note-41"><a href="#cite_ref-41">^</a> <a rel="nofollow" class="external text" href="https://www.image-net.org/static_files/files/ILSVRC2010_NEC-UIUC.pdf">ImageNet classification: fast descriptor coding and large-scale SVM training</a>

</li>

<li id="cite_note-42"><a href="#cite_ref-42">^</a> <cite id="CITEREFLinLvZhuYang2011" class="citation book cs1">Lin, Yuanqing; Lv, Fengjun; Zhu, Shenghuo; Yang, Ming; Cour, Timothee; Yu, Kai; Cao, Liangliang; Huang, Thomas (June 2011). <a rel="nofollow" class="external text" href="https://dx.doi.org/10.1109/cvpr.2011.5995477">"Large-scale image classification: Fast feature extraction and SVM training"</a>. CVPR 2011. IEEE. pp. 1689–1696. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1109%2Fcvpr.2011.5995477">10.1109/cvpr.2011.5995477</a>. <a href="ISBN_(identifier)" class="mw-redirect" title="ISBN (identifier)">ISBN</a> <bdi>978-1-4577-0394-2</bdi>.</cite>

</li>

<li id="cite_note-43"><a href="#cite_ref-43">^</a> <cite id="CITEREFSanchezPerronnin2011" class="citation book cs1">Sanchez, Jorge; Perronnin, Florent (June 2011). <a rel="nofollow" class="external text" href="https://dx.doi.org/10.1109/cvpr.2011.5995504">"High-dimensional signature compression for large-scale image classification"</a>. CVPR 2011. IEEE. pp. 1665–1672. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1109%2Fcvpr.2011.5995504">10.1109/cvpr.2011.5995504</a>. <a href="ISBN_(identifier)" class="mw-redirect" title="ISBN (identifier)">ISBN</a> <bdi>978-1-4577-0394-2</bdi>.</cite>

</li>

<li id="cite_note-44"><a href="#cite_ref-44">^</a> <cite id="CITEREFPerronninSánchezMensink2010" class="citation book cs1">Perronnin, Florent; Sánchez, Jorge; Mensink, Thomas (2010). <a rel="nofollow" class="external text" href="https://inria.hal.science/inria-00548630">"Improving the Fisher Kernel for Large-Scale Image Classification"</a>. In Daniilidis, Kostas; Maragos, Petros; Paragios, Nikos (eds.). Computer Vision – ECCV 2010. Lecture Notes in Computer Science. Vol. 6314. Berlin, Heidelberg: Springer. pp. 143–156. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1007%2F978-3-642-15561-1_11">10.1007/978-3-642-15561-1_11</a>. <a href="ISBN_(identifier)" class="mw-redirect" title="ISBN (identifier)">ISBN</a> <bdi>978-3-642-15561-1</bdi>.</cite>

</li>

<li id="cite_note-45"><a href="#cite_ref-45">^</a> <a rel="nofollow" class="external text" href="https://web.archive.org/web/20201027234359/http://image-net.org/challenges/LSVRC/2011/ilsvrc11.pdf">"XRCE@ILSVRC2011: Compressed Fisher vectors for LSVR"</a>, Florent Perronnin and Jorge Sánchez, Xerox Research Centre Europe (XRCE)

</li>

<li id="cite_note-46"><a href="#cite_ref-46">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://www.image-net.org/challenges/LSVRC/2012/results">"ImageNet Large Scale Visual Recognition Competition 2012 (ILSVRC2012)"</a>.</cite>

</li>

<li id="cite_note-47"><a href="#cite_ref-47">^</a> <cite id="CITEREFRussakovskyDengHuangBerg2013" class="citation journal cs1">Russakovsky, Olga; Deng, Jia; Huang, Zhiheng; Berg, Alexander C.; Fei-Fei, Li (2013). <a rel="nofollow" class="external text" href="https://openaccess.thecvf.com/content_iccv_2013/html/Russakovsky_Detecting_Avocados_to_2013_ICCV_paper.html">"Detecting Avocados to Zucchinis: What Have We Done, and Where Are We Going?"</a>: 2064–2071.</cite> <code class="cs1-code">{{cite journal}}</code>: Cite journal requires <code class="cs1-code">|journal=</code> (help)

</li>

<li id="cite_note-48"><a href="#cite_ref-48">^</a> <cite id="CITEREFSermanetEigenZhangMathieu2013" class="citation arxiv cs1">Sermanet, Pierre; Eigen, David; Zhang, Xiang; Mathieu, Michael; Fergus, Rob; LeCun, Yann (2013). "OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks". <a href="ArXiv_(identifier)" class="mw-redirect" title="ArXiv (identifier)">arXiv</a>:<a rel="nofollow" class="external text" href="https://arxiv.org/abs/1312.6229">1312.6229</a> [<a rel="nofollow" class="external text" href="https://arxiv.org/archive/cs.CV">cs.CV</a>].</cite>

</li>

<li id="cite_note-szegedy-49"><a href="#cite_ref-szegedy_49-0">^</a> <cite id="CITEREFSzegedyWei_LiuYangqing_JiaSermanet2015" class="citation book cs1">Szegedy, Christian; Wei Liu; Yangqing Jia; Sermanet, Pierre; Reed, Scott; Anguelov, Dragomir; Erhan, Dumitru; Vanhoucke, Vincent; Rabinovich, Andrew (June 2015). <a rel="nofollow" class="external text" href="https://ieeexplore.ieee.org/document/7298594">"Going deeper with convolutions"</a>. 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE. pp. 1–9. <a href="ArXiv_(identifier)" class="mw-redirect" title="ArXiv (identifier)">arXiv</a>:<a rel="nofollow" class="external text" href="https://arxiv.org/abs/1409.4842">1409.4842</a>. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1109%2FCVPR.2015.7298594">10.1109/CVPR.2015.7298594</a>. <a href="ISBN_(identifier)" class="mw-redirect" title="ISBN (identifier)">ISBN</a> <bdi>978-1-4673-6964-0</bdi>.</cite>

</li>

<li id="cite_note-50"><a href="#cite_ref-50">^</a> <cite id="CITEREFGershgorn2017" class="citation news cs1">Gershgorn, Dave (10 September 2017). <a rel="nofollow" class="external text" href="https://qz.com/1046350/the-quartz-guide-to-artificial-intelligence-what-is-it-why-is-it-important-and-should-we-be-afraid/">"The Quartz guide to artificial intelligence: What is it, why is it important, and should we be afraid?"</a>. Quartz. Retrieved 3 February 2018.</cite>

</li>

<li id="cite_note-51"><a href="#cite_ref-51">^</a> <cite id="CITEREFMarkoff2015" class="citation news cs1">Markoff, John (10 December 2015). <a rel="nofollow" class="external text" href="https://www.nytimes.com/2015/12/11/science/an-advance-in-artificial-intelligence-rivals-human-vision-abilities.html">"A Learning Advance in Artificial Intelligence Rivals Human Abilities"</a>. The New York Times. Retrieved 22 June 2016.</cite>

</li>

<li id="cite_note-52"><a href="#cite_ref-52">^</a> <cite id="CITEREFAron2015" class="citation news cs1">Aron, Jacob (21 September 2015). <a rel="nofollow" class="external text" href="https://www.newscientist.com/article/dn28206-forget-the-turing-test-there-are-better-ways-of-judging-ai/">"Forget the Turing test – there are better ways of judging AI"</a>. New Scientist. Retrieved 22 June 2016.</cite>

</li>

<li id="cite_note-53"><a href="#cite_ref-53">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://image-net.org/challenges/LSVRC/2016/results">"Ilsvrc2016"</a>.</cite>

</li>

<li id="cite_note-54"><a href="#cite_ref-54">^</a> <cite id="CITEREFXieGirshickDollarTu2017" class="citation conference cs1">Xie, Saining; Girshick, Ross; Dollar, Piotr; Tu, Zhuowen; <a href="Kaiming_He" title="Kaiming He">He, Kaiming</a> (2017). <a rel="nofollow" class="external text" href="https://openaccess.thecvf.com/content_cvpr_2017/papers/Xie_Aggregated_Residual_Transformations_CVPR_2017_paper.pdf">Aggregated Residual Transformations for Deep Neural Networks</a> (PDF). <a href="Conference_on_Computer_Vision_and_Pattern_Recognition" title="Conference on Computer Vision and Pattern Recognition">Conference on Computer Vision and Pattern Recognition</a>. pp. 1492–1500. <a href="ArXiv_(identifier)" class="mw-redirect" title="ArXiv (identifier)">arXiv</a>:<a rel="nofollow" class="external text" href="https://arxiv.org/abs/1611.05431">1611.05431</a>. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1109%2FCVPR.2017.634">10.1109/CVPR.2017.634</a>.</cite>

</li>

<li id="cite_note-55"><a href="#cite_ref-55">^</a> <cite id="CITEREFHuShenAlbanieSun2017" class="citation arxiv cs1">Hu, Jie; Shen, Li; Albanie, Samuel; Sun, Gang; Wu, Enhua (2017). "Squeeze-and-Excitation Networks". <a href="ArXiv_(identifier)" class="mw-redirect" title="ArXiv (identifier)">arXiv</a>:<a rel="nofollow" class="external text" href="https://arxiv.org/abs/1709.01507">1709.01507</a> [<a rel="nofollow" class="external text" href="https://arxiv.org/archive/cs.CV">cs.CV</a>].</cite>

</li>

<li id="cite_note-56"><a href="#cite_ref-56">^</a> <cite id="CITEREFNorthcuttAthalyeMueller2021" class="citation cs2">Northcutt, Curtis G.; Athalye, Anish; Mueller, Jonas (7 November 2021), Pervasive Label Errors in Test Sets Destabilize Machine Learning Benchmarks, <a href="ArXiv_(identifier)" class="mw-redirect" title="ArXiv (identifier)">arXiv</a>:<a rel="nofollow" class="external text" href="https://arxiv.org/abs/2103.14749">2103.14749</a></cite>

</li>

<li id="cite_note-57"><a href="#cite_ref-57">^</a> <cite id="CITEREFBeyerHénaffKolesnikovZhai2020" class="citation cs2">Beyer, Lucas; Hénaff, Olivier J.; Kolesnikov, Alexander; Zhai, Xiaohua; Oord, Aäron van den (12 June 2020), Are we done with ImageNet?, <a href="ArXiv_(identifier)" class="mw-redirect" title="ArXiv (identifier)">arXiv</a>:<a rel="nofollow" class="external text" href="https://arxiv.org/abs/2006.07159">2006.07159</a></cite>

</li>

<li id="cite_note-58"><a href="#cite_ref-58">^</a> <cite class="citation magazine cs1"><a rel="nofollow" class="external text" href="https://www.wired.com/story/viral-app-labels-you-isnt-what-you-think/">"The Viral App That Labels You Isn't Quite What You Think"</a>. Wired. <a href="ISSN_(identifier)" class="mw-redirect" title="ISSN (identifier)">ISSN</a> <a rel="nofollow" class="external text" href="https://search.worldcat.org/issn/1059-1028">1059-1028</a>. Retrieved 22 September 2019.</cite>

</li>

<li id="cite_note-59"><a href="#cite_ref-59">^</a> <cite id="CITEREFWong2019" class="citation news cs1"><a href="Julia_Carrie_Wong" title="Julia Carrie Wong">Wong, Julia Carrie</a> (18 September 2019). <a rel="nofollow" class="external text" href="https://www.theguardian.com/technology/2019/sep/17/imagenet-roulette-asian-racist-slur-selfie">"The viral selfie app ImageNet Roulette seemed fun – until it called me a racist slur"</a>. The Guardian. <a href="ISSN_(identifier)" class="mw-redirect" title="ISSN (identifier)">ISSN</a> <a rel="nofollow" class="external text" href="https://search.worldcat.org/issn/0261-3077">0261-3077</a>. Retrieved 22 September 2019.</cite>

</li>

<li id="cite_note-60"><a href="#cite_ref-60">^</a> <cite id="CITEREFCrawfordPaglen2019" class="citation web cs1">Crawford, Kate; Paglen, Trevor (19 September 2019). <a rel="nofollow" class="external text" href="https://www.excavating.ai/">"Excavating AI: The Politics of Training Sets for Machine Learning"</a>. -. Retrieved 22 September 2019.</cite>

</li>

<li id="cite_note-61"><a href="#cite_ref-61">^</a> <cite id="CITEREFLyons2020" class="citation journal cs1">Lyons, Michael (24 December 2020). "Excavating "Excavating AI": The Elephant in the Gallery". <a href="ArXiv_(identifier)" class="mw-redirect" title="ArXiv (identifier)">arXiv</a>:<a rel="nofollow" class="external text" href="https://arxiv.org/abs/2009.01215">2009.01215</a>. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.5281%2Fzenodo.4037538">10.5281/zenodo.4037538</a>.</cite> <code class="cs1-code">{{cite journal}}</code>: Cite journal requires <code class="cs1-code">|journal=</code> (help)

</li>

<li id="cite_note-62"><a href="#cite_ref-62">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="http://image-net.org/update-sep-17-2019.php">"Towards Fairer Datasets: Filtering and Balancing the Distribution of the People Subtree in the ImageNet Hierarchy"</a>. image-net.org. 17 September 2019. Retrieved 22 September 2019.</cite>

</li>

</ol></div>

<div class="mw-heading mw-heading3"><h3 id="Primary_sources">Primary sources</h3></div>

<ul><li><cite id="CITEREFDengDongSocherLi2009" class="citation conference cs1">Deng, Jia; Dong, Wei; Socher, Richard; Li, Li-Jia; Kai Li; Li Fei-Fei (June 2009). <a rel="nofollow" class="external text" href="https://ieeexplore.ieee.org/document/5206848">ImageNet: A large-scale hierarchical image database</a>. CVPR 2009. IEEE. pp. 248–255. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1109%2FCVPR.2009.5206848">10.1109/CVPR.2009.5206848</a>. <a href="ISBN_(identifier)" class="mw-redirect" title="ISBN (identifier)">ISBN</a> <bdi>978-1-4244-3992-8</bdi>.</cite></li>

<li><cite id="CITEREFFei-FeiDengLi2010" class="citation journal cs1">Fei-Fei, L.; Deng, J.; Li, K. (22 March 2010). <a rel="nofollow" class="external text" href="http://jov.arvojournals.org/Article.aspx?doi=10.1167/9.8.1037">"ImageNet: Constructing a large-scale image database"</a>. Journal of Vision. 9 (8): 1037. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1167%2F9.8.1037">10.1167/9.8.1037</a>. <a href="ISSN_(identifier)" class="mw-redirect" title="ISSN (identifier)">ISSN</a> <a rel="nofollow" class="external text" href="https://search.worldcat.org/issn/1534-7362">1534-7362</a>.</cite></li>

<li><cite id="CITEREFDengBergLiFei-Fei2010" class="citation conference cs1">Deng, Jia; Berg, Alexander C.; Li, Kai; Fei-Fei, Li (2010). Daniilidis, Kostas; Maragos, Petros; Paragios, Nikos (eds.). <a rel="nofollow" class="external text" href="https://link.springer.com/chapter/10.1007/978-3-642-15555-0_6">What Does Classifying More Than 10,000 Image Categories Tell Us?</a>. Computer Vision – ECCV 2010. Berlin, Heidelberg: Springer. pp. 71–84. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1007%2F978-3-642-15555-0_6">10.1007/978-3-642-15555-0_6</a>. <a href="ISBN_(identifier)" class="mw-redirect" title="ISBN (identifier)">ISBN</a> <bdi>978-3-642-15555-0</bdi>.</cite></li>

<li><cite id="CITEREFRussakovskyDengHuangBerg2013" class="citation conference cs1">Russakovsky, Olga; Deng, Jia; Huang, Zhiheng; Berg, Alexander C.; Fei-Fei, Li (2013). <a rel="nofollow" class="external text" href="https://openaccess.thecvf.com/content_iccv_2013/html/Russakovsky_Detecting_Avocados_to_2013_ICCV_paper.html">Detecting Avocados to Zucchinis: What Have We Done, and Where Are We Going?</a>. ICCV 2013. pp. 2064–2071.</cite></li>

<li><cite id="CITEREFRussakovskyDengSuKrause2015" class="citation journal cs1">Russakovsky, Olga; Deng, Jia; Su, Hao; Krause, Jonathan; Satheesh, Sanjeev; Ma, Sean; Huang, Zhiheng; Karpathy, Andrej; Khosla, Aditya; Bernstein, Michael; Berg, Alexander C.; Fei-Fei, Li (1 December 2015). <a rel="nofollow" class="external text" href="https://doi.org/10.1007/s11263-015-0816-y">"ImageNet Large Scale Visual Recognition Challenge"</a>. International Journal of Computer Vision. 115 (3): 211–252. <a href="ArXiv_(identifier)" class="mw-redirect" title="ArXiv (identifier)">arXiv</a>:<a rel="nofollow" class="external text" href="https://arxiv.org/abs/1409.0575">1409.0575</a>. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1007%2Fs11263-015-0816-y">10.1007/s11263-015-0816-y</a>. <a href="ISSN_(identifier)" class="mw-redirect" title="ISSN (identifier)">ISSN</a> <a rel="nofollow" class="external text" href="https://search.worldcat.org/issn/1573-1405">1573-1405</a>.</cite></li></ul>

<div class="mw-heading mw-heading2"><h2 id="External_links">External links</h2></div>

<ul><li><a rel="nofollow" class="external text" href="http://image-net.org">Official website</a></li></ul>

/* start https://en.wikipedia.org/ */

.mw-parser-output .hlist dl,.mw-parser-output .hlist ol,.mw-parser-output .hlist ul{margin:0;padding:0}.mw-parser-output .hlist dd,.mw-parser-output .hlist dt,.mw-parser-output .hlist li{margin:0;display:inline}.mw-parser-output .hlist.inline,.mw-parser-output .hlist.inline dl,.mw-parser-output .hlist.inline ol,.mw-parser-output .hlist.inline ul,.mw-parser-output .hlist dl dl,.mw-parser-output .hlist dl ol,.mw-parser-output .hlist dl ul,.mw-parser-output .hlist ol dl,.mw-parser-output .hlist ol ol,.mw-parser-output .hlist ol ul,.mw-parser-output .hlist ul dl,.mw-parser-output .hlist ul ol,.mw-parser-output .hlist ul ul{display:inline}.mw-parser-output .hlist .mw-empty-li{display:none}.mw-parser-output .hlist dt::after{content:": "}.mw-parser-output .hlist dd::after,.mw-parser-output .hlist li::after{content:" · ";font-weight:bold}.mw-parser-output .hlist dd:last-child::after,.mw-parser-output .hlist dt:last-child::after,.mw-parser-output .hlist li:last-child::after{content:none}.mw-parser-output .hlist dd dd:first-child::before,.mw-parser-output .hlist dd dt:first-child::before,.mw-parser-output .hlist dd li:first-child::before,.mw-parser-output .hlist dt dd:first-child::before,.mw-parser-output .hlist dt dt:first-child::before,.mw-parser-output .hlist dt li:first-child::before,.mw-parser-output .hlist li dd:first-child::before,.mw-parser-output .hlist li dt:first-child::before,.mw-parser-output .hlist li li:first-child::before{content:" (";font-weight:normal}.mw-parser-output .hlist dd dd:last-child::after,.mw-parser-output .hlist dd dt:last-child::after,.mw-parser-output .hlist dd li:last-child::after,.mw-parser-output .hlist dt dd:last-child::after,.mw-parser-output .hlist dt dt:last-child::after,.mw-parser-output .hlist dt li:last-child::after,.mw-parser-output .hlist li dd:last-child::after,.mw-parser-output .hlist li dt:last-child::after,.mw-parser-output .hlist li li:last-child::after{content:")";font-weight:normal}.mw-parser-output .hlist ol{counter-reset:listitem}.mw-parser-output .hlist ol>li{counter-increment:listitem}.mw-parser-output .hlist ol>li::before{content:" "counter(listitem)"\a0 "}.mw-parser-output .hlist dd ol>li:first-child::before,.mw-parser-output .hlist dt ol>li:first-child::before,.mw-parser-output .hlist li ol>li:first-child::before{content:" ("counter(listitem)"\a0 "}

/* end https://en.wikipedia.org/ */

</style><style data-mw-deduplicate="TemplateStyles:r1236075235">

/* start https://en.wikipedia.org/ */

.mw-parser-output .navbox{box-sizing:border-box;border:1px solid #a2a9b1;width:100%;clear:both;font-size:88%;text-align:center;padding:1px;margin:1em auto 0}.mw-parser-output .navbox .navbox{margin-top:0}.mw-parser-output .navbox+.navbox,.mw-parser-output .navbox+.navbox-styles+.navbox{margin-top:-1px}.mw-parser-output .navbox-inner,.mw-parser-output .navbox-subgroup{width:100%}.mw-parser-output .navbox-group,.mw-parser-output .navbox-title,.mw-parser-output .navbox-abovebelow{padding:0.25em 1em;line-height:1.5em;text-align:center}.mw-parser-output .navbox-group{white-space:nowrap;text-align:right}.mw-parser-output .navbox,.mw-parser-output .navbox-subgroup{background-color:#fdfdfd}.mw-parser-output .navbox-list{line-height:1.5em;border-color:#fdfdfd}.mw-parser-output .navbox-list-with-group{text-align:left;border-left-width:2px;border-left-style:solid}.mw-parser-output tr+tr>.navbox-abovebelow,.mw-parser-output tr+tr>.navbox-group,.mw-parser-output tr+tr>.navbox-image,.mw-parser-output tr+tr>.navbox-list{border-top:2px solid #fdfdfd}.mw-parser-output .navbox-title{background-color:#ccf}.mw-parser-output .navbox-abovebelow,.mw-parser-output .navbox-group,.mw-parser-output .navbox-subgroup .navbox-title{background-color:#ddf}.mw-parser-output .navbox-subgroup .navbox-group,.mw-parser-output .navbox-subgroup .navbox-abovebelow{background-color:#e6e6ff}.mw-parser-output .navbox-even{background-color:#f7f7f7}.mw-parser-output .navbox-odd{background-color:transparent}.mw-parser-output .navbox .hlist td dl,.mw-parser-output .navbox .hlist td ol,.mw-parser-output .navbox .hlist td ul,.mw-parser-output .navbox td.hlist dl,.mw-parser-output .navbox td.hlist ol,.mw-parser-output .navbox td.hlist ul{padding:0.125em 0}.mw-parser-output .navbox .navbar{display:block;font-size:100%}.mw-parser-output .navbox-title .navbar{float:left;text-align:left;margin-right:0.5em}body.skin--responsive .mw-parser-output .navbox-image img{max-width:none!important}@media print{body.ns-0 .mw-parser-output .navbox{display:none!important}}

/* end https://en.wikipedia.org/ */

</style></div><div role="navigation" class="navbox" aria-labelledby="Differentiable_computing254" style="padding:3px"><table class="nowraplinks hlist mw-collapsible autocollapse navbox-inner" style="border-spacing:0;background:transparent;color:inherit"><tbody><tr><th scope="col" class="navbox-title" colspan="2"><style data-mw-deduplicate="TemplateStyles:r1239400231">

/* start https://en.wikipedia.org/ */

.mw-parser-output .navbar{display:inline;font-size:88%;font-weight:normal}.mw-parser-output .navbar-collapse{float:left;text-align:left}.mw-parser-output .navbar-boxtext{word-spacing:0}.mw-parser-output .navbar ul{display:inline-block;white-space:nowrap;line-height:inherit}.mw-parser-output .navbar-brackets::before{margin-right:-0.125em;content:"[ "}.mw-parser-output .navbar-brackets::after{margin-left:-0.125em;content:" ]"}.mw-parser-output .navbar li{word-spacing:-0.125em}.mw-parser-output .navbar a>span,.mw-parser-output .navbar a>abbr{text-decoration:inherit}.mw-parser-output .navbar-mini abbr{font-variant:small-caps;border-bottom:none;text-decoration:none;cursor:inherit}.mw-parser-output .navbar-ct-full{font-size:114%;margin:0 7em}.mw-parser-output .navbar-ct-mini{font-size:114%;margin:0 4em}html.skin-theme-clientpref-night .mw-parser-output .navbar li a abbr{color:var(--color-base)!important}@media(prefers-color-scheme:dark){html.skin-theme-clientpref-os .mw-parser-output .navbar li a abbr{color:var(--color-base)!important}}@media print{.mw-parser-output .navbar{display:none!important}}

/* end https://en.wikipedia.org/ */

</style><div id="Differentiable_computing254" style="font-size:114%;margin:0 4em">Differentiable computing</div></th></tr><tr><th scope="row" class="navbox-group" style="width:1%"><a href="Differentiable_function" title="Differentiable function">General</a></th><td class="navbox-list-with-group navbox-list navbox-odd" style="width:100%;padding:0"><div style="padding:0 0.25em">

<ul><li><a href="Differentiable_programming" title="Differentiable programming">Differentiable programming</a></li>

<li><a href="Information_geometry" title="Information geometry">Information geometry</a></li>

<li><a href="Statistical_manifold" title="Statistical manifold">Statistical manifold</a></li>

<li><a href="Automatic_differentiation" title="Automatic differentiation">Automatic differentiation</a></li>

<li><a href="Neuromorphic_computing" title="Neuromorphic computing">Neuromorphic computing</a></li>

<li><a href="Pattern_recognition" title="Pattern recognition">Pattern recognition</a></li>

<li><a href="Ricci_calculus" title="Ricci calculus">Ricci calculus</a></li>

<li><a href="Computational_learning_theory" title="Computational learning theory">Computational learning theory</a></li>

<li><a href="Inductive_bias" title="Inductive bias">Inductive bias</a></li></ul>

</div></td></tr><tr><th scope="row" class="navbox-group" style="width:1%">Hardware</th><td class="navbox-list-with-group navbox-list navbox-even" style="width:100%;padding:0"><div style="padding:0 0.25em">

<li><a href="Memristor" title="Memristor">Memristor</a></li>

<li><a href="SpiNNaker" title="SpiNNaker">SpiNNaker</a></li></ul>

</div></td></tr><tr><th scope="row" class="navbox-group" style="width:1%">Software libraries</th><td class="navbox-list-with-group navbox-list navbox-odd" style="width:100%;padding:0"><div style="padding:0 0.25em">

<ul><li><a href="TensorFlow" title="TensorFlow">TensorFlow</a></li>

<li><a href="PyTorch" title="PyTorch">PyTorch</a></li>

<li><a href="Keras" title="Keras">Keras</a></li>

<li><a href="Scikit-learn" title="Scikit-learn">scikit-learn</a></li>

<li><a href="Theano_(software)" title="Theano (software)">Theano</a></li>

<li><a href="MindSpore" title="MindSpore">MindSpore</a></li></ul>

</div></td></tr><tr><td class="navbox-abovebelow" colspan="2"><div>

<ul><li> Portals

<ul><li><a href="Portal%3AComputer_programming" title="Portal:Computer programming">Computer programming</a></li>

<li><a href="Portal%3ATechnology" title="Portal:Technology">Technology</a></li></ul></li></ul>

</div></td></tr></tbody></table></div>

<div class="navbox-styles"></div><div role="navigation" class="navbox" aria-labelledby="Standard_test_items314" style="padding:3px"><table class="nowraplinks mw-collapsible autocollapse navbox-inner" style="border-spacing:0;background:transparent;color:inherit"><tbody><tr><th scope="col" class="navbox-title" colspan="2"><div id="Standard_test_items314" style="font-size:114%;margin:0 4em">Standard test items</div></th></tr><tr><td class="navbox-abovebelow hlist" colspan="2"><div>

<ul><li><a href="Pangram" title="Pangram">Pangram</a></li>

<li><a href="Reference_implementation" title="Reference implementation">Reference implementation</a></li>

<li><a href="Sanity_check" title="Sanity check">Sanity check</a></li>

<li><a href="Standard_test_image" title="Standard test image">Standard test image</a></li></ul>

</div></td></tr><tr><th scope="row" class="navbox-group" style="width:1%"><a href="Artificial_intelligence" title="Artificial intelligence">Artificial intelligence</a> (<a href="Machine_learning" title="Machine learning">Machine learning</a>)</th><td class="navbox-list-with-group navbox-list navbox-odd hlist" style="width:100%;padding:0"><div style="padding:0 0.25em">

<ul><li><a href="Chinese_room" title="Chinese room">Chinese room</a></li>

<li><a href="MNIST_database" title="MNIST database">MNIST database</a></li>

<li><a href="Turing_test" title="Turing test">Turing test</a></li>

</div></td></tr><tr><th scope="row" class="navbox-group" style="width:1%">Television (<a href="Test_card" title="Test card">test card</a>)</th><td class="navbox-list-with-group navbox-list navbox-even hlist" style="width:100%;padding:0"><div style="padding:0 0.25em">

<ul><li><a href="SMPTE_color_bars" title="SMPTE color bars">SMPTE color bars</a></li>

<li><a href="EBU_colour_bars" title="EBU colour bars">EBU colour bars</a></li>

<li><a href="Indian-head_test_pattern" title="Indian-head test pattern">Indian-head test pattern</a></li>

<li><a href="EIA_1956_resolution_chart" title="EIA 1956 resolution chart">EIA 1956 resolution chart</a></li>

<li><a href="Philips_circle_pattern" title="Philips circle pattern">Philips circle pattern</a> (<a href="Philips_circle_pattern#PM5534" title="Philips circle pattern">PM 5538</a>, <a href="Philips_PM5540" title="Philips PM5540">PM 5540</a>, <a href="Philips_circle_pattern#PM5544" title="Philips circle pattern">PM 5544</a>, <a href="Philips_circle_pattern#PM5644" title="Philips circle pattern">PM 5644</a>)</li>

<li><a href="Snell_%26_Wilcox_Zone_Plate" title="Snell & Wilcox Zone Plate">Snell & Wilcox SW2/SW4</a></li>

<li><a href="Telefunken_FuBK" title="Telefunken FuBK">Telefunken FuBK</a></li>

</div></td></tr><tr><th scope="row" class="navbox-group" style="width:1%"><a href="Computer_language" title="Computer language">Computer languages</a></th><td class="navbox-list-with-group navbox-list navbox-odd hlist" style="width:100%;padding:0"><div style="padding:0 0.25em">

<ul><li><a href="%22Hello%2C_World!%22_program" title=""Hello, World!" program">"Hello, World!" program</a></li>

<li><a href="Quine_(computing)" title="Quine (computing)">Quine</a></li>

<li><a href="TPK_algorithm" title="TPK algorithm">Trabb Pardo–Knuth algorithm</a></li>

<li><a href="Perl#Community" title="Perl">Just another Perl hacker</a></li></ul>

</div></td></tr><tr><th scope="row" class="navbox-group" style="width:1%"><a href="Data_compression" title="Data compression">Data compression</a></th><td class="navbox-list-with-group navbox-list navbox-even hlist" style="width:100%;padding:0"><div style="padding:0 0.25em">

<ul><li><a href="Calgary_corpus" title="Calgary corpus">Calgary corpus</a></li>

<li><a href="Canterbury_corpus" title="Canterbury corpus">Canterbury corpus</a></li>

<li><a href="Silesia_corpus" title="Silesia corpus">Silesia corpus</a></li>

<li><a href="Hutter_Prize" title="Hutter Prize">enwik8, enwik9</a></li></ul>

</div></td></tr><tr><th scope="row" class="navbox-group" style="width:1%"><a href="3D_computer_graphics" title="3D computer graphics">3D computer graphics</a></th><td class="navbox-list-with-group navbox-list navbox-odd hlist" style="width:100%;padding:0"><div style="padding:0 0.25em">

<ul><li><a href="3DBenchy" title="3DBenchy">3DBenchy</a></li>

<li><a href="Cornell_box" title="Cornell box">Cornell box</a></li>

<li><a href="Stanford_bunny" title="Stanford bunny">Stanford bunny</a></li>

<li><a href="Stanford_dragon" title="Stanford dragon">Stanford dragon</a></li>

<li><a href="Utah_teapot" title="Utah teapot">Utah teapot</a></li>

</div></td></tr><tr><th scope="row" class="navbox-group" style="width:1%"><a href="2D_computer_graphics" title="2D computer graphics">2D computer graphics</a></th><td class="navbox-list-with-group navbox-list navbox-even hlist" style="width:100%;padding:0"><div style="padding:0 0.25em">

<ul><li>Ghostscript tiger</li>

</div></td></tr><tr><th scope="row" class="navbox-group" style="width:1%"><a href="Typography" title="Typography">Typography</a> (<a href="Filler_text" title="Filler text">filler text</a>)</th><td class="navbox-list-with-group navbox-list navbox-odd hlist" style="width:100%;padding:0"><div style="padding:0 0.25em">

<ul><li><a href="Etaoin_shrdlu" title="Etaoin shrdlu">Etaoin shrdlu</a></li>

<li><a href="Hamburgevons" title="Hamburgevons">Hamburgevons</a></li>

<li><a href="Lorem_ipsum" title="Lorem ipsum">Lorem ipsum</a></li>

<li><a href="The_quick_brown_fox_jumps_over_the_lazy_dog" title="The quick brown fox jumps over the lazy dog">The quick brown fox jumps over the lazy dog</a></li></ul>

</div></td></tr><tr><th scope="row" class="navbox-group" style="width:1%">Other</th><td class="navbox-list-with-group navbox-list navbox-even hlist" style="width:100%;padding:0"><div style="padding:0 0.25em">

<ul><li>Acid

<li><a href="Bad_Apple!!#Use_of_video_as_a_graphical_and_audio_test" title="Bad Apple!!">"Bad Apple!!"</a></li>

<li><a href="EICAR_test_file" title="EICAR test file">EICAR test file</a></li>

<li><a href="Test_functions_for_optimization" title="Test functions for optimization">Functions for optimization</a></li>

<li><a href="GTUBE" title="GTUBE">GTUBE</a></li>

<li><a href="Harvard_sentences" title="Harvard sentences">Harvard sentences</a></li>

<li><a href="The_North_Wind_and_the_Sun#Use_in_phonetic_demonstrations" title="The North Wind and the Sun">"The North Wind and the Sun"</a></li>

<li><a href="Tom's_Diner#The_"Mother_of_the_MP3"" title="Tom's Diner">"Tom's Diner"</a></li>

<li><a href="Film_leader" title="Film leader">SMPTE universal leader</a></li>

<li><a href="EURion_constellation" title="EURion constellation">EURion constellation</a></li>

<li><a href="Webdriver_Torso" title="Webdriver Torso">Webdriver Torso</a></li>

<li><a href="1951_USAF_resolution_test_chart" title="1951 USAF resolution test chart">1951 USAF resolution test chart</a></li></ul>

</div></td></tr></tbody></table></div></div><div><div class="zim-footer">

This article is issued from <a class="external text" title="Last edited on 2025-07-28" href="https://en.wikipedia.org/wiki/?title=ImageNet&oldid=1302992656">Wikipedia</a>. The text is available under <a class="external text" href="https://creativecommons.org/licenses/by-sa/4.0/deed.en">Creative Commons Attribution-Share Alike 4.0</a> unless otherwise noted. Additional terms may apply for the media files.

</div>

</div></div>

</div>

</main>

</div>

</body></html>