Datasets

5,457 machine learning datasets
🔔   Share your dataset with the ML community!
Filter by Modality
Images 1684 Texts 1516 Videos 540 Audio 253 Medical 195 3D 161 Graphs 125 Speech 119 RGB-D 97 Environment 81 Time series 68 Point cloud 48 Tabular 46 Biomedical 45 LiDAR 29 Biology 23 Stereo 22 Tracking 22 Hyperspectral images 20 MRI 17 Interactive 15 3d meshes 14 Physics 14 Actions 13 Music 13 RGB Video 13 Dialog 12 Tables 12 EEG 9 Midi 7 6D 6 Ranking 5 Cad 4 Financial 4 Parallel 4 Replay data 4 fMRI 3 Lyrics 2 PSG 2
Filter by Task
Question Answering 265 Semantic Segmentation 205 Object Detection 169 Image Classification 146 Language Modelling 123 Reading Comprehension 89 Named Entity Recognition 82 Pose Estimation 77 Action Recognition 75 Text Classification 73 Visual Question Answering 71 Sentiment Analysis 69 Text Generation 66 Domain Adaptation 64 Speech Recognition 62 Data Augmentation 59 Information Retrieval 58 Machine Translation 58 Natural Language Inference 58 Autonomous Driving 54 Text Summarization 54 Depth Estimation 52 Image Generation 52 Natural Language Understanding 50 Relation Extraction 47 Word Embeddings 47 Image Captioning 46 Link Prediction 46 Face Recognition 43 Anomaly Detection 42 Image Retrieval 42 Instance Segmentation 41 Person Re-Identification 41 Multi-Task Learning 40 Coreference Resolution 37 Machine Reading Comprehension 37 Temporal Action Localization 37 Scene Understanding 36 Decision Making 34 Object Recognition 34 Recommendation Systems 34 Semantic Parsing 34 Common Sense Reasoning 33 Abstractive Text Summarization 32 Knowledge Graphs 32 Medical Image Segmentation 32 Object Tracking 32 Graph Classification 31 Image Super-Resolution 29 Emotion Recognition 28 Fine-Grained Image Classification 28 Node Classification 28 Optical Character Recognition 28 Optical Flow Estimation 28 Image-to-Image Translation 26 Video Understanding 26 3D Reconstruction 25 Document Summarization 25 Metric Learning 25 Neural Architecture Search 25 Autonomous Vehicles 24 Face Detection 24 Misinformation 24 Multi-Label Classification 23 Self-Supervised Learning 23 Entity Linking 22 Super-Resolution 22 Trajectory Prediction 22 3D Human Pose Estimation 21 3D Object Detection 21 Action Detection 21 Unsupervised Domain Adaptation 21 Video Captioning 21 Visual Reasoning 21 Activity Recognition 20 Fake News Detection 20 Few-Shot Learning 20 Scene Classification 20 Video Retrieval 20 Action Classification 19 Face Verification 19 Skeleton Based Action Recognition 19 3D Pose Estimation 18 Domain Generalization 18 Facial Expression Recognition 18 Image Clustering 18 Zero-Shot Learning 18 Code Generation 17 Data-to-Text Generation 17 Dialogue Generation 17 Hand Pose Estimation 17 Hate Speech Detection 17 Music Information Retrieval 17 Open-Domain Question Answering 17 Part-Of-Speech Tagging 17 Question Generation 17 Video Question Answering 17 Face Alignment 16 Few-Shot Image Classification 16 Graph Clustering 16 Imitation Learning 16 Meta-Learning 16 Music Generation 16 Scene Text Recognition 16 Sound Event Detection 16 Visual Object Tracking 16 Visual Tracking 16 Word Sense Disambiguation 16 3D Semantic Segmentation 15 Continual Learning 15 Cross-Lingual Transfer 15 Crowd Counting 15 Handwriting Recognition 15 Monocular Depth Estimation 15 Paraphrase Identification 15 Semantic Textual Similarity 15 Style Transfer 15 Video Prediction 15 Audio Classification 14 Density Estimation 14 Emotion Classification 14 Fairness 14 Human-Object Interaction Detection 14 Multi-Object Tracking 14 Salient Object Detection 14 Speech Synthesis 14 Visual Localization 14 Visual Odometry 14 2D Semantic Segmentation 13 3D Hand Pose Estimation 13 Active Learning 13 Computed Tomography (CT) 13 Denoising 13 Face Anti-Spoofing 13 Fact Verification 13 Multi-Document Summarization 13 Object Counting 13 Relation Classification 13 Sign Language Recognition 13 Slot Filling 13 Speech Enhancement 13 Stereo Matching 13 Vision and Language Navigation 13 Visual Navigation 13 Audio Source Separation 12 Cross-Modal Retrieval 12 Gesture Recognition 12 Image Dehazing 12 Image Inpainting 12 Intent Detection 12 Panoptic Segmentation 12 Pedestrian Detection 12 Quantization 12 RGB Salient Object Detection 12 Saliency Detection 12 Scene Recognition 12 Scene Text Detection 12 Stochastic Optimization 12 Unsupervised Anomaly Detection 12 Video Classification 12 Clustering Algorithms Evaluation 11 Document Classification 11 Hand Gesture Recognition 11 Language Identification 11 Out-of-Distribution Detection 11 Paraphrase Generation 11 Referring Expression Segmentation 11 Robotic Grasping 11 Self-Driving Cars 11 Sentence Embeddings 11 Text Simplification 11 Activity Detection 10 Age Estimation 10 Aspect-Based Sentiment Analysis 10 COVID-19 Diagnosis 10 Deblurring 10 DeepFake Detection 10 Dependency Parsing 10 Dimensionality Reduction 10 Facial Landmark Detection 10 Gaze Estimation 10 Human Detection 10 Image Compression 10 Multiple Object Tracking 10 Object Localization 10 Real-Time Semantic Segmentation 10 Relational Reasoning 10 Semi-Supervised Image Classification 10 Sentence Classification 10 Simultaneous Localization and Mapping 10 Small Data Image Classification 10 Stance Detection 10 Video Summarization 10 Visual Dialog 10 Weakly Supervised Object Detection 10 Weather Forecasting 10 2D Human Pose Estimation 9 Action Recognition In Videos 9 Animal Pose Estimation 9 Automatic Post-Editing 9 Code Search 9 Conversational Response Selection 9 Entity Disambiguation 9 Lane Detection 9 Learning-To-Rank 9 Lesion Segmentation 9 Nested Named Entity Recognition 9 Novel View Synthesis 9 Outlier Detection 9 Person Search 9 Trajectory Forecasting 9 Transfer Learning 9 3D Face Reconstruction 8 3D Instance Segmentation 8 Acoustic Scene Classification 8 Dialogue State Tracking 8 Emotion Recognition in Conversation 8 Event Extraction 8 Generalizable Person Re-identification 8 Generalized Zero-Shot Learning 8 Grammatical Error Correction 8 Graph Embedding 8 Handwriting generation 8 Human Part Segmentation 8 Image Denoising 8 Image Enhancement 8 Image Restoration 8 Image/Document Clustering 8 Knowledge Graph Completion 8 Long-tail Learning 8 Low-Light Image Enhancement 8 Motion Estimation 8 Multi-Label Text Classification 8 Real-Time Object Detection 8 Spoken Language Understanding 8 Task-Oriented Dialogue Systems 8 Text-To-Sql 8 Time Series Forecasting 8 Token Classification 8 Vehicle Re-Identification 8 Video Generation 8 Visual Place Recognition 8 motion prediction 8 3D Action Recognition 7 6D Pose Estimation 7 Abusive Language 7 Action Segmentation 7 Automated Theorem Proving 7 Colorectal Polyps Characterization 7 Community Detection 7 Continuous Control 7 Depth Completion 7 Disaster Response 7 Document Layout Analysis 7 Edge Detection 7 Entity Resolution 7 Face Swapping 7 Federated Learning 7 Graph Regression 7 Homography Estimation 7 Human action generation 7 Imputation 7 Incremental Learning 7 Intent Classification 7 Keypoint Detection 7 Knowledge Base Question Answering 7 Learning with noisy labels 7 Lesion Classification 7 Multiple Instance Learning 7 Open Information Extraction 7 Opinion Mining 7 Point Cloud Registration 7 Robot Navigation 7 Semantic Similarity 7 Sequence-to-sequence Language Modeling 7 Single Image Dehazing 7 Source Code Summarization 7 Table Detection 7 Unsupervised Video Object Segmentation 7 Video Frame Interpolation 7 Video Quality Assessment 7 Video Super-Resolution 7 3D Object Recognition 6 Adversarial Robustness 6 Answer Selection 6 Atari Games 6 Audio Generation 6 Brain Tumor Segmentation 6 Chatbot 6 Chinese Reading Comprehension 6 Code Summarization 6 Color Image Denoising 6 Constituency Parsing 6 Curriculum Learning 6 Dictionary Learning 6 Discourse Parsing 6 Document Ranking 6 Entity Typing 6 Goal-Oriented Dialog 6 Graph Matching 6 Hyperspectral Image Classification 6 Image Matting 6 Interactive Segmentation 6 KG-to-Text Generation 6 License Plate Recognition 6 Multi-class Classification 6 Multimodal Activity Recognition 6 Multivariate Time Series Forecasting 6 Music Modeling 6 Music Source Separation 6 Node Clustering 6 Offline RL 6 Program Repair 6 RGB-D Salient Object Detection 6 Sarcasm Detection 6 Scientific Document Summarization 6 Semantic Role Labeling 6 Sentence Embedding 6 Sign Language Translation 6 Speaker Diarization 6 Speech Separation 6 Starcraft 6 Table-to-Text Generation 6 Text-Image Retrieval 6 Text-To-Speech Synthesis 6 Time Series 6 Unsupervised Image Classification 6 Unsupervised Person Re-Identification 6 Video Description 6 Video Inpainting 6 Video Object Detection 6 Video Object Segmentation 6 Video-Based Person Re-Identification 6 Virtual Try-on 6 3D Human Reconstruction 5 6D Pose Estimation using RGB 5 Action Quality Assessment 5 Adversarial Attack 5 Adversarial Defense 5 Audio Tagging 5 AutoML 5 Binarization 5 Boundary Detection 5 Change Point Detection 5 Chinese Named Entity Recognition 5 Class-agnostic Object Detection 5 Code Documentation Generation 5 Colorization 5 Conditional Image Generation 5 Content-Based Image Retrieval 5 Contrastive Learning 5 Core set discovery 5 Dense Video Captioning 5 Dialogue Understanding 5 Drug Discovery 5 EEG 5 Explainable artificial intelligence 5 Face Identification 5 Face Model 5 Face Presentation Attack Detection 5 Face Sketch Synthesis 5 Feature Engineering 5 Fine-Grained Visual Categorization 5 Fine-Grained Visual Recognition 5 Font Recognition 5 Food Recognition 5 Graph Generation 5 Handwritten Text Recognition 5 Head Pose Estimation 5 Human Interaction Recognition 5 Image Quality Assessment 5 Image Registration 5 JPEG Artifact Correction 5 Joint Entity and Relation Extraction 5 Keyword Spotting 5 Language Acquisition 5 Lexical Entailment 5 Lipreading 5 Math Word Problem Solving 5 Mathematical Question Answering 5 Medical Named Entity Recognition 5 Medical Relation Extraction 5 Morphological Analysis 5 Motion Segmentation 5 Multi-Person Pose Estimation 5 Multimodal Sentiment Analysis 5 Music Classification 5 Music Transcription 5 Network Pruning 5 News Classification 5 Object Detection In Aerial Images 5 Object Proposal Generation 5 Online Multi-Object Tracking 5 Partial Domain Adaptation 5 Pedestrian Attribute Recognition 5 Product Recommendation 5 Region Proposal 5 Scene Segmentation 5 Semantic correspondence 5 Semi-Supervised Video Object Segmentation 5 Spatio-Temporal Action Localization 5 Speaker Recognition 5 Stereo Matching Hand 5 Structured Prediction 5 Surface Normals Estimation 5 Text Matching 5 Text-to-Image Generation 5 Time Series Analysis 5 Time Series Classification 5 Topic Models 5 Traffic Prediction 5 Tumor Segmentation 5 Unsupervised Object Segmentation 5 Video Recognition 5 Video Segmentation 5 Video Semantic Segmentation 5 Vision-Language Navigation 5 Weakly-Supervised Named Entity Recognition 5 Zero-Shot Transfer Image Classification 5 3D Face Animation 4 3D Feature Matching 4 3D Medical Imaging Segmentation 4 3D Multi-Object Tracking 4 3D Multi-Person Pose Estimation 4 3D Object Tracking 4 3D Point Cloud Classification 4 Action Anticipation 4 Aesthetics Quality Assessment 4 Autonomous Navigation 4 Bayesian Inference 4 Bias Detection 4 Brain Segmentation 4 Causal Inference 4 Classification with Binary Weight Network 4 Click-Through Rate Prediction 4 Clinical Concept Extraction 4 Color Constancy 4 Community Question Answering 4 Compressive Sensing 4 Cross-Lingual NER 4 Cross-Lingual Question Answering 4 Defect Detection 4 Dialogue Act Classification 4 Discourse Segmentation 4 Egocentric Activity Recognition 4 Environmental Sound Classification 4 Event-based vision 4 Extractive Text Summarization 4 Fact Checking 4 Feature Importance 4 Few-Shot Object Detection 4 Fine-Grained Image Recognition 4 Gaussian Processes 4 General Classification 4 General Reinforcement Learning 4 Genre classification 4 Goal-Oriented Dialogue Systems 4 Graph Representation Learning 4 Grasp Contact Prediction 4 Hand Segmentation 4 Hand-Gesture Recognition 4 Human Pose Forecasting 4 Human robot interaction 4 Image Manipulation 4 Image Stitching 4 Image-based Automatic Meter Reading 4 Keyword Extraction 4 LIDAR Semantic Segmentation 4 License Plate Detection 4 Line Segment Detection 4 Low Resource Named Entity Recognition 4 Matrix Completion 4 Medical Diagnosis 4 Meter Reading 4 Monocular 3D Object Detection 4 Multi-Label Learning 4 Multi-Source Unsupervised Domain Adaptation 4 Multi-agent Reinforcement Learning 4 Multi-future Trajectory Prediction 4 Multimodal Deep Learning 4 Multimodal Machine Translation 4 Multimodal Unsupervised Image-To-Image Translation 4 NER 4 Natural Language Visual Grounding 4 Nested Mention Recognition 4 Network Embedding 4 Network Intrusion Detection 4 Open World Object Detection 4 Parameter Prediction 4 Passage Retrieval 4 Patch Matching 4 Physical Simulations 4 Probabilistic Deep Learning 4 Real-Time Strategy Games 4 Rectification 4 Referring Expression Comprehension 4 Remote Sensing Image Classification 4 Retinal Vessel Segmentation 4 Robust Object Detection 4 Saliency Prediction 4
Filter by Language
English 1254 Chinese 166 German 115 French 84 Russian 76 Spanish 75 Japanese 56 Arabic 54 Italian 53 Portuguese 48 Turkish 41 Korean 36 Hindi 35 Dutch 31 Vietnamese 30 Persian 27 Polish 25 Czech 24 Finnish 24 Tamil 24 Romanian 23 Bengali 21 Indonesian 21 Multilingual 19 Telugu 19 Urdu 19 Thai 17 Basque 15 Danish 15 Estonian 15 Malayalam 15 Mandarin Chinese 15 Marathi 15 Swedish 15 Hungarian 14 Bulgarian 13 Kannada 13 Gujarati 12 Catalan 11 Hebrew 11 Norwegian 11 Ukrainian 11 Greek 10 Kazakh 10 Latvian 10 Punjabi 10 Slovak 10 Slovenian 10 Amharic 9 Croatian 9 Serbian 9 Swahili 9 Welsh 9 Albanian 8 Armenian 8 Assamese 8 Breton 8 Lithuanian 8 Sinhala 8 Georgian 7 Mongolian 7 Esperanto 6 Icelandic 6 Irish 6 Kurdish 6 Macedonian 6 Maltese 6 Oriya (macrolanguage) 6 Sanskrit 6 Yoruba 6 Afrikaans 5 American Sign Language 5 Azerbaijani 5 Galician 5 Igbo 5 Latin 5 Scottish Gaelic 5 Sindhi 5 Tagalog 5 Uzbek 5 Belarusian 4 Bosnian 4 Burmese 4 Chechen 4 Haitian 4 Hausa 4 Javanese 4 Malagasy 4 Nepali (macrolanguage) 4 Serbo-Croatian 4 Somali 4 Standard Arabic 4 Sundanese 4 Tatar 4 Upper Sorbian 4 Wolof 4 Aragonese 3 Bambara 3 Bavarian 3 Bishnupriya 3 Central Khmer 3 Chuvash 3 Dhivehi 3 Egyptian Arabic 3 Erzya 3 Filipino 3 Guarani 3 Iranian Persian 3 Kinyarwanda 3 Kirghiz 3 Lao 3 Malay (individual language) 3 Norwegian Nynorsk 3 Oromo 3 Quechua 3 Romansh 3 Russia Buriat 3 South Azerbaijani 3 Swiss German 3 Tibetan 3 Uighur 3 Yiddish 3 Asturian 2 Avaric 2 Bashkir 2 Cebuano 2 Central Bikol 2 Central Kurdish 2 Cherokee 2 Church Slavic 2 Cornish 2 Dimli (individual language) 2 Eastern Mari 2 Faroese 2 Fon 2 Fulah 2 Ganda 2 Goan Konkani 2 Gothic 2 Ido 2 Iloko 2 Interlingue 2 Inuktitut 2 Jejueo 2 Kabyle 2 Kalmyk 2 Karachay-Balkar 2 Komi 2 Komi-Permyak 2 Lezghian 2 Limburgan 2 Lingala 2 Livvi 2 Lojban 2 Lombard 2 Low German 2 Lower Sorbian 2 Luxembourgish 2 Maithili 2 Manipuri 2 Manx 2 Mazanderani 2 Minangkabau 2 Mingrelian 2 Mirandese 2 Modern Greek 2 Moksha 2 Naxi 2 Neapolitan 2 Newari 2 Nigerian Pidgin 2 Northern Frisian 2 Northern Luri 2 Northern Sami 2 Occitan (post 1500) 2 Odia 2 Ossetian 2 Pampanga 2 Piemontese 2 Pushto 2 Sardinian 2 Sichuan Yi 2 Sicilian 2 Swati 2 Tai 2 Tajik 2 Tigrinya 2 Tswana 2 Turkish Sign Language 2 Turkmen 2 Tuvinian 2 Venetian 2 Volapük 2 Walloon 2 Waray (Philippines) 2 Western Frisian 2 Western Mari 2 Western Panjabi 2 Wu Chinese 2 Xhosa 2 Yakut 2 Yue Chinese 2 Abkhazian 1 Achinese 1 Adyghe 1 Afar 1 Akan 1 Akkadian 1 Akuntsu 1 Ancient Greek 1 Apurinã 1 Argentine Sign Language 1 Arpitan 1 Assyrian Neo-Aramaic 1 Aymara 1 Bangladeshi Sign Language 1 Banjar 1 Bhojpuri 1 Bislama 1 Bodo (India) 1 Buginese 1 Central Pashto 1 Chamorro 1 Chavacano 1 Cheyenne 1 Choctaw 1 Chukot 1 Coptic 1 Corsican 1 Cree 1 Creek 1 Crimean Tatar 1 Dzongkha 1 Ewe 1 Extremaduran 1 Fiji Hindi 1 Fijian 1 Friulian 1 Gagauz 1 Gan Chinese 1 Geez 1 German Sign Language 1 Gilaki 1 Greek Sign Language 1 Gulf Arabic 1 Hakha Chin 1 Hakka Chinese 1 Hawaiian 1 Herero 1 Hiri Motu 1 Interlingua (International Auxiliary Language Association) 1 Inupiaq 1 Jamaican Creole English 1 Kabardian 1 Kalaallisut 1 Kanuri 1 Kara-Kalpak 1 Karelian 1 Kashmiri 1 Kashubian 1 Khunsari 1 Kikuyu 1 Komi-Zyrian 1 Kongo 1 Kuanyama 1 Kölsch 1 Ladino 1 Lak 1 Latgalian 1 Ligurian 1 Literary Chinese 1 Luo (Cameroon) 1 Luo (Kenya and Tanzania) 1 Malay (macrolanguage) 1 Maori 1 Marshallese 1 Mbyá Guaraní 1 Min Dong Chinese 1 Modern Greek (1453-) 1 Moroccan Arabic 1 Mundurukú 1 Narom 1 Nauru 1 Navajo 1 Nayini 1 Ndonga 1 Nepali (individual language) 1 Northern Kurdish 1 Norwegian Bokmål 1 Novial 1 Nyanja 1 Official Aramaic (700-300 BCE) 1 Old English (ca. 450-1100) 1 Old French 1 Old Russian 1 Old Turkish 1 Pali 1 Pangasinan 1 Papiamento 1 Pedi 1 Pennsylvania German 1 Pfaelzisch 1 Picard 1 Pitcairn-Norfolk 1 Pontic 1 Portuguse 1 Rajasthani 1 Rundi 1 Rusyn 1 Samoan 1 Sango 1 Saterfriesisch 1 Scots 1 Shona 1 Silesian 1 Skolt Sami 1 Soi 1 South Levantine Arabic 1 Southern Sotho 1 Sranan Tongo 1 Swahili (macrolanguage) 1 Swedish Sign Language 1 Swiss-German Sign Language 1 Tahitian 1 Tetum 1 Tok Pisin 1 Tonga (Tonga Islands) 1 Tosk Albanian 1 Tsonga 1 Tulu 1 Tumbuka 1 Tunisian Arabic 1 Tupinambá 1 Twi 1 Udmurt 1 Venda 1 Veps 1 Vlaams 1 Vlax Romani 1 Votic 1 Warlpiri 1 Zeeuws 1 Zhuang 1 Zulu 1 Northern Huishui Hmong 0 Santali 0

5457 dataset results

fastMRI

The fastMRI dataset includes two types of MRI scans: knee MRIs and the brain (neuro) MRIs, and containing training, validation, and masked test sets. The deidentified imaging dataset provided by NYU Langone comprises raw k-space data in several sub-dataset groups. Curation of these data are part of an IRB approved study. Raw and DICOM data have been deidentified via conversion to the vendor-neutral ISMRMD format and the RSNA clinical trial processor, respectively. Also, each DICOM image is manually inspected for the presence of any unexpected protected health information (PHI), with spot checking of both metadata and image content. Knee MRI: Data from more than 1,500 fully sampled knee MRIs obtained on 3 and 1.5 Tesla magnets and DICOM images from 10,000 clinical knee MRIs also obtained at 3 or 1.5 Tesla. The raw dataset includes coronal proton density-weighted images with and without fat suppression. The DICOM dataset contains coronal proton density-weighted with and without fat suppr

117 PAPERS • 4 BENCHMARKS

CIFAR-FS (CIFAR100 few-shots)

CIFAR100 few-shots (CIFAR-FS) is randomly sampled from CIFAR-100 (Krizhevsky & Hinton, 2009) by using the same criteria with which miniImageNet has been generated. The average inter-class similarity is sufficiently high to represent a challenge for the current state of the art. Moreover, the limited original resolution of 32×32 makes the task harder and at the same time allows fast prototyping.

116 PAPERS • 2 BENCHMARKS

MOTChallenge

The MOTChallenge datasets are designed for the task of multiple object tracking. There are several variants of the dataset released each year, such as MOT15, MOT17, MOT20.

116 PAPERS • 7 BENCHMARKS

VizDoom

ViZDoom is an AI research platform based on the classical First Person Shooter game Doom. The most popular game mode is probably the so-called Death Match, where several players join in a maze and fight against each other. After a fixed time, the match ends and all the players are ranked by the FRAG scores defined as kills minus suicides. During the game, each player can access various observations, including the first-person view screen pixels, the corresponding depth-map and segmentation-map (pixel-wise object labels), the bird-view maze map, etc. The valid actions include almost all the keyboard-stroke and mouse-control a human player can take, accounting for moving, turning, jumping, shooting, changing weapon, etc. ViZDoom can run a game either synchronously or asynchronously, indicating whether the game core waits until all players’ actions are collected or runs in a constant frame rate without waiting.

116 PAPERS • 2 BENCHMARKS

Billion Word Benchmark
Billion Word Benchmark

The One Billion Word dataset is a dataset for language modeling. The training/held-out data was produced from the WMT 2011 News Crawl data using a combination of Bash shell and Perl scripts.

115 PAPERS • 2 BENCHMARKS

100DOH (100 Days Of Hands Dataset)

The 100 Days Of Hands Dataset (100DOH) is a large-scale video dataset containing hands and hand-object interactions. It consists of 27.3K Youtube videos from 11 categories with nearly 131 days of footage of everyday interaction. The focus of the dataset is hand contact, and it includes both first-person and third-person perspectives. The videos in 100DOH are unconstrained and content-rich, ranging from records of daily life to specific instructional videos. To enforce diversity, the dataset contains no more than 20 videos from each uploader.

113 PAPERS • 2 BENCHMARKS

DOTA (Dataset for Object deTection in Aerial Images)

DOTA is a large-scale dataset for object detection in aerial images. It can be used to develop and evaluate object detectors in aerial images. The images are collected from different sensors and platforms. Each image is of the size in the range from 800 × 800 to 20,000 × 20,000 pixels and contains objects exhibiting a wide variety of scales, orientations, and shapes. The instances in DOTA images are annotated by experts in aerial image interpretation by arbitrary (8 d.o.f.) quadrilateral. We will continue to update DOTA, to grow in size and scope to reflect evolving real-world conditions. Now it has three versions:

112 PAPERS • 4 BENCHMARKS

NELL (Never Ending Language Learning)

NELL is a dataset built from the Web via an intelligent agent called Never-Ending Language Learner. This agent attempts to learn over time to read the web. NELL has accumulated over 50 million candidate beliefs by reading the web, and it is considering these at different levels of confidence. NELL has high confidence in 2,810,379 of these beliefs.

112 PAPERS • 1 BENCHMARK

ENZYMES
ENZYMES

ENZYMES is a dataset of 600 protein tertiary structures obtained from the BRENDA enzyme database. The ENZYMES dataset contains 6 enzymes.

111 PAPERS • 1 BENCHMARK

MOT17 (Multiple Object Tracking 17)

The Multiple Object Tracking 17 (MOT17) dataset is a dataset for multiple object tracking. Similar to its previous version MOT16, this challenge contains seven different indoor and outdoor scenes of public places with pedestrians as the objects of interest. A video for each scene is divided into two clips, one for training and the other for testing. The dataset provides detections of objects in the video frames with three detectors, namely SDP, Faster-RCNN and DPM. The challenge accepts both on-line and off-line tracking approaches, where the latter are allowed to use the future video frames to predict tracks.

110 PAPERS • 2 BENCHMARKS

ROCStories

ROCStories is a collection of commonsense short stories. The corpus consists of 100,000 five-sentence stories. Each story logically follows everyday topics created by Amazon Mechanical Turk workers. These stories contain a variety of commonsense causal and temporal relations between everyday events. Writers also develop an additional 3,742 Story Cloze Test stories which contain a four-sentence-long body and two candidate endings. The endings were collected by asking Mechanical Turk workers to write both a right ending and a wrong ending after eliminating original endings of given short stories. Both endings were required to make logical sense and include at least one character from the main story line. The published ROCStories dataset is constructed with ROCStories as a training set that includes 98,162 stories that exclude candidate wrong endings, an evaluation set, and a test set, which have the same structure (1 body + 2 candidate endings) and a size of 1,871.

110 PAPERS • 2 BENCHMARKS

Something-Something V2 (20BN-Something-Something Dataset V2)

The 20BN-SOMETHING-SOMETHING V2 dataset is a large collection of labeled video clips that show humans performing pre-defined basic actions with everyday objects. The dataset was created by a large number of crowd workers. It allows machine learning models to develop fine-grained understanding of basic actions that occur in the physical world. It contains 220,847 videos, with 168,913 in the training set, 24,777 in the validation set and 27,157 in the test set. There are 174 labels.

110 PAPERS • 3 BENCHMARKS

Stanford Online Products

Stanford Online Products (SOP) dataset has 22,634 classes with 120,053 product images. The first 11,318 classes (59,551 images) are split for training and the other 11,316 (60,502 images) classes are used for testing

110 PAPERS • 1 BENCHMARK

VisDA-2017

VisDA-2017 is a simulation-to-real dataset for domain adaptation with over 280,000 images across 12 categories in the training, validation and testing domains. The training images are generated from the same object under different circumstances, while the validation images are collected from MSCOCO..

110 PAPERS • 3 BENCHMARKS

ActivityNet Captions

The ActivityNet Captions dataset is built on ActivityNet v1.3 which includes 20k YouTube untrimmed videos with 100k caption annotations. The videos are 120 seconds long on average. Most of the videos contain over 3 annotated events with corresponding start/end time and human-written sentences, which contain 13.5 words on average. The number of videos in train/validation/test split is 10024/4926/5044, respectively.

108 PAPERS • 4 BENCHMARKS

RefCOCO

This referring expression generation (REG) dataset was collected using the ReferitGame. In this two-player game, the first player is shown an image with a segmented target object and asked to write a natural language expression referring to the target object. The second player is shown only the image and the referring expression and asked to click on the corresponding object. If the players do their job correctly, they receive points and swap roles. If not, they are presented with a new object and image for description. Images in these collections were selected to contain two or more objects of the same object category. In the RefCOCO dataset, no restrictions are placed on the type of language used in the referring expressions. In a version of this dataset called RefCOCO+ players are disallowed from using location words in their referring expressions by adding “taboo” words to the ReferItGame. This dataset was collected to obtain a referring expression dataset focsed on purely appearan

108 PAPERS • 8 BENCHMARKS

BoolQ (Boolean Questions)

BoolQ is a question answering dataset for yes/no questions containing 15942 examples. These questions are naturally occurring – they are generated in unprompted and unconstrained settings. Each example is a triplet of (question, passage, answer), with the title of the page as optional additional context.

107 PAPERS • 1 BENCHMARK

IAM (IAM Handwriting)

The IAM database contains 13,353 images of handwritten lines of text created by 657 writers. The texts those writers transcribed are from the Lancaster-Oslo/Bergen Corpus of British English. It includes contributions from 657 writers making a total of 1,539 handwritten pages comprising of 115,320 words and is categorized as part of modern collection. The database is labeled at the sentence, line, and word levels.

107 PAPERS • 1 BENCHMARK

TUM RGB-D

TUM RGB-D is an RGB-D dataset. It contains the color and depth images of a Microsoft Kinect sensor along the ground-truth trajectory of the sensor. The data was recorded at full frame rate (30 Hz) and sensor resolution (640x480). The ground-truth trajectory was obtained from a high-accuracy motion-capture system with eight high-speed tracking cameras (100 Hz).

107 PAPERS • NO BENCHMARKS YET

UCF-QNRF

The UCF-QNRF dataset is a crowd counting dataset and it contains large diversity both in scenes, as well as in background types. It consists of 1535 images high-resolution images from Flickr, Web Search and Hajj footage. The number of people (i.e., the count) varies from 50 to 12,000 across images.

107 PAPERS • 1 BENCHMARK

COPA (Choice of Plausible Alternatives)

The Choice Of Plausible Alternatives (COPA) evaluation provides researchers with a tool for assessing progress in open-domain commonsense causal reasoning. COPA consists of 1000 questions, split equally into development and test sets of 500 questions each. Each question is composed of a premise and two alternatives, where the task is to select the alternative that more plausibly has a causal relation with the premise. The correct alternative is randomized so that the expected performance of randomly guessing is 50%.

106 PAPERS • 1 BENCHMARK

Foggy Cityscapes

Foggy Cityscapes is a synthetic foggy dataset which simulates fog on real scenes. Each foggy image is rendered with a clear image and depth map from Cityscapes. Thus the annotations and data split in Foggy Cityscapes are inherited from Cityscapes.

106 PAPERS • 3 BENCHMARKS

MOT16 (Multiple Object Tracking 2016)

The MOT16 dataset is a dataset for multiple object tracking. It a collection of existing and new data (part of the sources are from and ), containing 14 challenging real-world videos of both static scenes and moving scenes, 7 for training and 7 for testing. It is a large-scale dataset, composed of totally 110407 bounding boxes in training set and 182326 bounding boxes in test set. All video sequences are annotated under strict standards, their ground-truths are highly accurate, making the evaluation meaningful.

106 PAPERS • 2 BENCHMARKS

OTB-2013

OTB2013 is the previous version of the current OTB2015 Visual Tracker Benchmark. It contains only 50 tracking sequences, as opposed to the 100 sequences in the current version of the benchmark.

106 PAPERS • 2 BENCHMARKS

VOT2018

VOT2018 is a dataset for visual object tracking. It consists of 60 challenging videos collected from real-life datasets.

106 PAPERS • 1 BENCHMARK

Chairs

The Chairs dataset contains rendered images of around 1000 different three-dimensional chair models.

105 PAPERS • 1 BENCHMARK

GOT-10k (Generic Object Tracking Benchmark)

The GOT-10k dataset contains more than 10,000 video segments of real-world moving objects and over 1.5 million manually labelled bounding boxes. The dataset contains more than 560 classes of real-world moving objects and 80+ classes of motion patterns.

105 PAPERS • 1 BENCHMARK

ORL (Our Database of Faces)

The ORL Database of Faces contains 400 images from 40 distinct subjects. For some subjects, the images were taken at different times, varying the lighting, facial expressions (open / closed eyes, smiling / not smiling) and facial details (glasses / no glasses). All the images were taken against a dark homogeneous background with the subjects in an upright, frontal position (with tolerance for some side movement). The size of each image is 92x112 pixels, with 256 grey levels per pixel.

105 PAPERS • 1 BENCHMARK

CORD-19

CORD-19 is a free resource of tens of thousands of scholarly articles about COVID-19, SARS-CoV-2, and related coronaviruses for use by the global research community.

104 PAPERS • NO BENCHMARKS YET

HowTo100M

HowTo100M is a large-scale dataset of narrated videos with an emphasis on instructional videos where content creators teach complex tasks with an explicit intention of explaining the visual content on screen. HowTo100M features a total of:

104 PAPERS • 1 BENCHMARK

YouTube-8M

The YouTube-8M dataset is a large scale video dataset, which includes more than 7 million videos with 4716 classes labeled by the annotation system. The dataset consists of three parts: training set, validate set, and test set. In the training set, each class contains at least 100 training videos. Features of these videos are extracted by the state-of-the-art popular pre-trained models and released for public use. Each video contains audio and visual modality. Based on the visual information, videos are divided into 24 topics, such as sports, game, arts & entertainment, etc

104 PAPERS • 2 BENCHMARKS

D4RL

D4RL is a collection of environments for offline reinforcement learning. These environments include Maze2D, AntMaze, Adroit, Gym, Flow, FrankKitchen and CARLA.

102 PAPERS • NO BENCHMARKS YET

SNAP (Standford Large Network Dataset Collection)

SNAP is a collection of large network datasets. It includes graphs representing social networks, citation networks, web graphs, online communities, online reviews and more.

102 PAPERS • NO BENCHMARKS YET

UCY

The UCY dataset consist of real pedestrian trajectories with rich multi-human interaction scenarios captured at 2.5 Hz (Δt=0.4s). It is composed of three sequences (Zara01, Zara02, and UCY), taken in public spaces from top-view.

102 PAPERS • NO BENCHMARKS YET

VOT2016

VOT2016 is a video dataset for visual object tracking. It contains 60 video clips and 21,646 corresponding ground truth maps with pixel-wise annotation of salient objects.

102 PAPERS • 1 BENCHMARK

ZINC

ZINC is a free database of commercially-available compounds for virtual screening. ZINC contains over 230 million purchasable compounds in ready-to-dock, 3D formats. ZINC also contains over 750 million purchasable compounds that can be searched for analogs.

102 PAPERS • 5 BENCHMARKS

LJSpeech (The LJ Speech Dataset)

This is a public domain speech dataset consisting of 13,100 short audio clips of a single speaker reading passages from 7 non-fiction books. A transcription is provided for each clip. Clips vary in length from 1 to 10 seconds and have a total length of approximately 24 hours. The texts were published between 1884 and 1964, and are in the public domain. The audio was recorded in 2016-17 by the LibriVox project and is also in the public domain.

101 PAPERS • 3 BENCHMARKS

SURREAL (Synthetic Humans for REAL Tasks)

SURREAL (Synthetic hUmans foR REAL tasks) is a large-scale person dataset that generates photorealistic synthetic images with labeling for human part segmentation and depth estimation, producing 6.5M frames in 67.5K short clips (about 100 frames each) of 2.6K action sequences with 145 different synthetic subjects. To ensure realism, the synthetic bodies are created using the SMPL body model, whose parameters are fit by the MoSh method given raw 3D MoCap marker data.

101 PAPERS • 1 BENCHMARK

UTKFace

The UTKFace dataset is a large-scale face dataset with long age span (range from 0 to 116 years old). The dataset consists of over 20,000 face images with annotations of age, gender, and ethnicity. The images cover large variation in pose, facial expression, illumination, occlusion, resolution, etc. This dataset could be used on a variety of tasks, e.g., face detection, age estimation, age progression/regression, landmark localization, etc.

101 PAPERS • 1 BENCHMARK

LRW (Lip Reading in the Wild)

The Lip Reading in the Wild (LRW) dataset a large-scale audio-visual database that contains 500 different words from over 1,000 speakers. Each utterance has 29 frames, whose boundary is centered around the target word. The database is divided into training, validation and test sets. The training set contains at least 800 utterances for each class while the validation and test sets contain 50 utterances.

99 PAPERS • 6 BENCHMARKS

Make3D

The Make3D dataset is a monocular Depth Estimation dataset that contains 400 single training RGB and depth map pairs, and 134 test samples. The RGB images have high resolution, while the depth maps are provided at low resolution.

99 PAPERS • 1 BENCHMARK

QuAC (Question Answering in Context)

Question Answering in Context is a large-scale dataset that consists of around 14K crowdsourced Question Answering dialogs with 98K question-answer pairs in total. Data instances consist of an interactive dialog between two crowd workers: (1) a student who poses a sequence of freeform questions to learn as much as possible about a hidden Wikipedia text, and (2) a teacher who answers the questions by providing short excerpts (spans) from the text.

98 PAPERS • 1 BENCHMARK

SMAC
SMAC (The StarCraft Multi-Agent Challenge)

The StarCraft Multi-Agent Challenge (SMAC) is a benchmark that provides elements of partial observability, challenging dynamics, and high-dimensional observation spaces. SMAC is built using the StarCraft II game engine, creating a testbed for research in cooperative MARL where each game unit is an independent RL agent.

98 PAPERS • 5 BENCHMARKS

Common Voice

Common Voice is an audio dataset that consists of a unique MP3 and corresponding text file. There are 9,283 recorded hours in the dataset. The dataset also includes demographic metadata like age, sex, and accent. The dataset consists of 7,335 validated hours in 60 languages.

97 PAPERS • 62 BENCHMARKS

LVIS

LVIS is a dataset for long tail instance segmentation. It has annotations for over 1000 object categories in 164k images.

97 PAPERS • 5 BENCHMARKS

AI2-THOR

AI2-Thor is an interactive environment for embodied AI. It contains four types of scenes, including kitchen, living room, bedroom and bathroom, and each scene includes 30 rooms, where each room is unique in terms of furniture placement and item types. There are over 2000 unique objects for AI agents to interact with.

96 PAPERS • 1 BENCHMARK

TrackingNet

TrackingNet is a large-scale tracking dataset consisting of videos in the wild. It has a total of 30,643 videos split into 30,132 training videos and 511 testing videos, with an average of 470,9 frames.

96 PAPERS • 1 BENCHMARK

WebVision

The WebVision dataset is designed to facilitate the research on learning visual representation from noisy web data. It is a large scale web images dataset that contains more than 2.4 million of images crawled from the Flickr website and Google Images search.

96 PAPERS • 2 BENCHMARKS

Contact us on: hello@paperswithcode.com . Papers With Code is a free resource with all data licensed under CC-BY-SA.
Terms Data policy Cookies policy