The MS COCO (Microsoft Common Objects in Context) dataset is a large-scale object detection, segmentation, key-point detection, and captioning dataset. The dataset consists of 328K images.

9,210 PAPERS • 87 BENCHMARKS

UCF101 dataset is an extension of UCF50 and consists of 13,320 video clips, which are classified into 101 categories. These 101 categories can be classified into 5 types (Body motion, Human-human interactions, Human-object interactions, Playing musical instruments and Sports). The total length of these video clips is over 27 hours. All the videos are collected from YouTube and have a fixed frame rate of 25 FPS with the resolution of 320 × 240.

1,542 PAPERS • 23 BENCHMARKS

The Kinetics dataset is a large-scale, high-quality dataset for human action recognition in videos. The dataset consists of around 500,000 video clips covering 600 human action classes with at least 600 video clips for each action class. Each video clip lasts around 10 seconds and is labeled with a single action class. The videos are collected from YouTube.

1,074 PAPERS • 20 BENCHMARKS

The HMDB51 dataset is a large collection of realistic videos from various sources, including movies and web videos. The dataset is composed of 6,766 video clips from 51 action categories (such as “jump”, “kiss” and “laugh”), with each category containing at least 101 clips. The original evaluation scheme uses three different training/testing splits. In each split, each action class has 70 clips for training and 30 clips for testing. The average accuracy over these three splits is used to measure the final performance.

739 PAPERS • 18 BENCHMARKS

The Human3.6M dataset is one of the largest motion capture datasets, which consists of 3.6 million human poses and corresponding images captured by a high-speed motion capture system. There are 4 high-resolution progressive scan cameras to acquire video data at 50 Hz. The dataset contains activities by 11 professional actors in 17 scenarios: discussion, smoking, taking photo, talking on the phone, etc., as well as provides accurate 3D joint positions and high-resolution videos.

671 PAPERS • 16 BENCHMARKS

The ActivityNet dataset contains 200 different types of activities and a total of 849 hours of videos collected from YouTube. ActivityNet is the largest benchmark for temporal activity detection to date in terms of both the number of activity categories and number of videos, making the task particularly challenging. Version 1.3 of the dataset contains 19994 untrimmed videos in total and is divided into three disjoint subsets, training, validation, and testing by a ratio of 2:1:1. On average, each activity category has 137 untrimmed videos. Each video on average has 1.41 activities which are annotated with temporal boundaries. The ground-truth annotations of test videos are not public.

629 PAPERS • 15 BENCHMARKS

The Densely Annotation Video Segmentation dataset (DAVIS) is a high quality and high resolution densely annotated video segmentation dataset under two resolutions, 480p and 1080p. There are 50 video sequences with 3455 densely annotated frames in pixel level. 30 videos with 2079 frames are for training and 20 videos with 1376 frames are for validation.

591 PAPERS • 13 BENCHMARKS

Multimodal Emotion Recognition IEMOCAP The IEMOCAP dataset consists of 151 videos of recorded dialogues, with 2 speakers per session for a total of 302 videos across the dataset. Each segment is annotated for the presence of 9 emotions (angry, excited, fear, sad, surprised, frustrated, happy, disappointed and neutral) as well as valence, arousal and dominance. The dataset is recorded across 5 sessions with 5 pairs of speakers.

562 PAPERS • 3 BENCHMARKS

MSR-VTT (Microsoft Research Video to Text) is a large-scale dataset for the open domain video captioning, which consists of 10,000 video clips from 20 categories, and each video clip is annotated with 20 English sentences by Amazon Mechanical Turks. There are about 29,000 unique words in all captions. The standard splits uses 6,513 clips for training, 497 clips for validation, and 2,990 clips for testing.

487 PAPERS • 9 BENCHMARKS

VoxCeleb2 is a large scale speaker recognition dataset obtained automatically from open-source media. VoxCeleb2 consists of over a million utterances from over 6k speakers. Since the dataset is collected ‘in the wild’, the speech segments are corrupted with real world noise including laughter, cross-talk, channel effects, music and other sounds. The dataset is also multilingual, with speech from speakers of 145 different nationalities, covering a wide range of accents, ages, ethnicities and languages. The dataset is audio-visual, so is also useful for a number of other applications, for example – visual speech synthesis, speech separation, cross-modal transfer from face to voice or vice versa and training face recognition from video to complement existing face recognition datasets.

452 PAPERS • 5 BENCHMARKS

Object Tracking Benchmark (OTB) is a visual tracking benchmark that is widely used to evaluate the performance of a visual tracking algorithm. The dataset contains a total of 100 sequences and each is annotated frame-by-frame with bounding boxes and 11 challenge attributes. OTB-2013 dataset contains 51 sequences and the OTB-2015 dataset contains all 100 sequences of the OTB dataset.

383 PAPERS • 4 BENCHMARKS

NTU RGB+D is a large-scale dataset for RGB-D human action recognition. It involves 56,880 samples of 60 action classes collected from 40 subjects. The actions can be generally divided into three categories: 40 daily actions (e.g., drinking, eating, reading), nine health-related actions (e.g., sneezing, staggering, falling down), and 11 mutual actions (e.g., punching, kicking, hugging). These actions take place under 17 different scene conditions corresponding to 17 video sequences (i.e., S001–S017). The actions were captured using three cameras with different horizontal imaging viewpoints, namely, −45∘,0∘, and +45∘. Multi-modality information is provided for action characterization, including depth maps, 3D skeleton joint position, RGB frames, and infrared sequences. The performance evaluation is performed by a cross-subject test that split the 40 subjects into training and test groups, and by a cross-view test that employed one camera (+45∘) for testing, and the other two cameras for

369 PAPERS • 5 BENCHMARKS

The Charades dataset is composed of 9,848 videos of daily indoors activities with an average length of 30 seconds, involving interactions with 46 objects classes in 15 types of indoor scenes and containing a vocabulary of 30 verbs leading to 157 action classes. Each video in this dataset is annotated by multiple free-text descriptions, action labels, action intervals and classes of interacting objects. 267 different users were presented with a sentence, which includes objects and actions from a fixed vocabulary, and they recorded a video acting out the sentence. In total, the dataset contains 66,500 temporal annotations for 157 action classes, 41,104 labels for 46 object classes, and 27,847 textual descriptions of the videos. In the standard split there are7,986 training video and 1,863 validation video.

355 PAPERS • 6 BENCHMARKS

The 3D Poses in the Wild dataset is the first dataset in the wild with accurate 3D poses for evaluation. While other datasets outdoors exist, they are all restricted to a small recording volume. 3DPW is the first one that includes video footage taken from a moving phone camera.

299 PAPERS • 5 BENCHMARKS

Datasets drive vision progress, yet existing driving datasets are impoverished in terms of visual content and supported tasks to study multitask learning for autonomous driving. Researchers are usually constrained to study a small set of problems on one dataset, while real-world computer vision applications require performing tasks of various complexities. We construct BDD100K, the largest driving video dataset with 100K videos and 10 tasks to evaluate the exciting progress of image recognition algorithms on autonomous driving. The dataset possesses geographic, environmental, and weather diversity, which is useful for training models that are less likely to be surprised by new conditions. Based on this diverse dataset, we build a benchmark for heterogeneous multitask learning and study how to solve the tasks together. Our experiments show that special training strategies are needed for existing models to perform such heterogeneous tasks. BDD100K opens the door for future studies in thi

298 PAPERS • 15 BENCHMARKS

The Microsoft Research Video Description Corpus (MSVD) dataset consists of about 120K sentences collected during the summer of 2010. Workers on Mechanical Turk were paid to watch a short video snippet and then summarize the action in a single sentence. The result is a set of roughly parallel descriptions of more than 2,000 video snippets. Because the workers were urged to complete the task in the language of their choice, both paraphrase and bilingual alternations are captured in the data.

285 PAPERS • 5 BENCHMARKS

FaceForensics++ is a forensics dataset consisting of 1000 original video sequences that have been manipulated with four automated face manipulation methods: Deepfakes, Face2Face, FaceSwap and NeuralTextures. The data has been sourced from 977 youtube videos and all videos contain a trackable mostly frontal face without occlusions which enables automated tampering methods to generate realistic forgeries.

284 PAPERS • 2 BENCHMARKS

Argoverse is a tracking benchmark with over 30K scenarios collected in Pittsburgh and Miami. Each scenario is a sequence of frames sampled at 10 HZ. Each sequence has an interesting object called “agent”, and the task is to predict the future locations of agents in a 3 seconds future horizon. The sequences are split into training, validation and test sets, which have 205,942, 39,472 and 78,143 sequences respectively. These splits have no geographical overlap.

283 PAPERS • 6 BENCHMARKS

The THUMOS14 dataset is a large-scale video dataset that includes 1,010 videos for validation and 1,574 videos for testing from 20 classes. Among all the videos, there are 220 and 212 videos with temporal annotations in validation and testing set, respectively.

280 PAPERS • 20 BENCHMARKS

The GoPro dataset for deblurring consists of 3,214 blurred images with the size of 1,280×720 that are divided into 2,103 training images and 1,111 test images. The dataset consists of pairs of a realistic blurry image and the corresponding ground truth shapr image that are obtained by a high-speed camera.

267 PAPERS • 3 BENCHMARKS

DAVIS17 is a dataset for video object segmentation. It contains a total of 150 videos - 60 for training, 30 for validation, 60 for testing

249 PAPERS • 11 BENCHMARKS

The efforts to create a non-trivial and publicly available dataset for action recognition was initiated at the KTH Royal Institute of Technology in 2004. The KTH dataset is one of the most standard datasets, which contains six actions: walk, jog, run, box, hand-wave, and hand clap. To account for performance nuance, each action is performed by 25 different individuals, and the setting is systematically altered for each action per actor. Setting variations include: outdoor (s1), outdoor with scale variation (s2), outdoor with different clothes (s3), and indoor (s4). These variations test the ability of each algorithm to identify actions independent of the background, appearance of the actors, and the scale of the actors.

248 PAPERS • 1 BENCHMARK

MPI-INF-3DHP is a 3D human body pose estimation dataset consisting of both constrained indoor and complex outdoor scenes. It records 8 actors performing 8 activities from 14 camera views. It consists on >1.3M frames captured from the 14 cameras.

237 PAPERS • 6 BENCHMARKS

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.

227 PAPERS • 1 BENCHMARK

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:

225 PAPERS • 1 BENCHMARK

JHMDB is an action recognition dataset that consists of 960 video sequences belonging to 21 actions. It is a subset of the larger HMDB51 dataset collected from digitized movies and YouTube videos. The dataset contains video and annotation for puppet flow per frame (approximated optimal flow on the person), puppet mask per frame, joint positions per frame, action label per clip and meta label per clip (camera motion, visible body parts, camera viewpoint, number of people, video quality).

223 PAPERS • 8 BENCHMARKS

MSMT17 is a multi-scene multi-time person re-identification dataset. The dataset consists of 180 hours of videos, captured by 12 outdoor cameras, 3 indoor cameras, and during 12 time slots. The videos cover a long period of time and present complex lighting variations, and it contains a large number of annotated identities, i.e., 4,101 identities and 126,441 bounding boxes.

218 PAPERS • 6 BENCHMARKS

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.

218 PAPERS • 8 BENCHMARKS

LaSOT is a high-quality benchmark for Large-scale Single Object Tracking. LaSOT consists of 1,400 sequences with more than 3.5M frames in total. Each frame in these sequences is carefully and manually annotated with a bounding box, making LaSOT one of the largest densely annotated tracking benchmark. The average video length of LaSOT is more than 2,500 frames, and each sequence comprises various challenges deriving from the wild where target objects may disappear and re-appear again in the view.

213 PAPERS • 2 BENCHMARKS

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.

211 PAPERS • NO BENCHMARKS YET

DAVIS16 is a dataset for video object segmentation which consists of 50 videos in total (30 videos for training and 20 for testing). Per-frame pixel-wise annotations are offered.

209 PAPERS • 4 BENCHMARKS

The Extended Cohn-Kanade (CK+) dataset contains 593 video sequences from a total of 123 different subjects, ranging from 18 to 50 years of age with a variety of genders and heritage. Each video shows a facial shift from the neutral expression to a targeted peak expression, recorded at 30 frames per second (FPS) with a resolution of either 640x490 or 640x480 pixels. Out of these videos, 327 are labelled with one of seven expression classes: anger, contempt, disgust, fear, happiness, sadness, and surprise. The CK+ database is widely regarded as the most extensively used laboratory-controlled facial expression classification database available, and is used in the majority of facial expression classification methods.

208 PAPERS • 2 BENCHMARKS

CamVid (Cambridge-driving Labeled Video Database) is a road/driving scene understanding database which was originally captured as five video sequences with a 960×720 resolution camera mounted on the dashboard of a car. Those sequences were sampled (four of them at 1 fps and one at 15 fps) adding up to 701 frames. Those stills were manually annotated with 32 classes: void, building, wall, tree, vegetation, fence, sidewalk, parking block, column/pole, traffic cone, bridge, sign, miscellaneous text, traffic light, sky, tunnel, archway, road, road shoulder, lane markings (driving), lane markings (non-driving), animal, pedestrian, child, cart luggage, bicyclist, motorcycle, car, SUV/pickup/truck, truck/bus, train, and other moving object

206 PAPERS • 3 BENCHMARKS

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.

201 PAPERS • 5 BENCHMARKS

The IJB-C dataset is a video-based face recognition dataset. It is an extension of the IJB-A dataset with about 138,000 face images, 11,000 face videos, and 10,000 non-face images.

199 PAPERS • 3 BENCHMARKS

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.

189 PAPERS • 2 BENCHMARKS

Multimodal EmotionLines Dataset (MELD) has been created by enhancing and extending EmotionLines dataset. MELD contains the same dialogue instances available in EmotionLines, but it also encompasses audio and visual modality along with text. MELD has more than 1400 dialogues and 13000 utterances from Friends TV series. Multiple speakers participated in the dialogues. Each utterance in a dialogue has been labeled by any of these seven emotions -- Anger, Disgust, Sadness, Joy, Neutral, Surprise and Fear. MELD also has sentiment (positive, negative and neutral) annotation for each utterance.

180 PAPERS • 2 BENCHMARKS

The Moving MNIST dataset contains 10,000 video sequences, each consisting of 20 frames. In each video sequence, two digits move independently around the frame, which has a spatial resolution of 64×64 pixels. The digits frequently intersect with each other and bounce off the edges of the frame

176 PAPERS • 1 BENCHMARK

The Vimeo-90K is a large-scale high-quality video dataset for lower-level video processing. It proposes three different video processing tasks: frame interpolation, video denoising/deblocking, and video super-resolution.

176 PAPERS • 3 BENCHMARKS

OTB-2015, also referred as Visual Tracker Benchmark, is a visual tracking dataset. It contains 100 commonly used video sequences for evaluating visual tracking. Image Source: http://cvlab.hanyang.ac.kr/tracker_benchmark/datasets.html

171 PAPERS • 1 BENCHMARK

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.

169 PAPERS • 2 BENCHMARKS

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.

167 PAPERS • 7 BENCHMARKS

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.

167 PAPERS • 4 BENCHMARKS

MARS (Motion Analysis and Re-identification Set) is a large scale video based person reidentification dataset, an extension of the Market-1501 dataset. It has been collected from six near-synchronized cameras. It consists of 1,261 different pedestrians, who are captured by at least 2 cameras. The variations in poses, colors and illuminations of pedestrians, as well as the poor image quality, make it very difficult to yield high matching accuracy. Moreover, the dataset contains 3,248 distractors in order to make it more realistic. Deformable Part Model and GMMCP tracker were used to automatically generate the tracklets (mostly 25-50 frames long).

164 PAPERS • 2 BENCHMARKS

The Sports-1M dataset consists of over a million videos from YouTube. The videos in the dataset can be obtained through the YouTube URL specified by the authors. Approximately 7% (as of 2016) of the videos have been removed by the YouTube uploaders since the dataset was compiled. However, there are still over a million videos in the dataset with 487 sports-related categories with 1,000 to 3,000 videos per category. The videos are automatically labelled with 487 sports classes using the YouTube Topics API by analyzing the text metadata associated with the videos (e.g. tags, descriptions). Approximately 5% of the videos are annotated with more than one class.

160 PAPERS • 2 BENCHMARKS

Charades-STA is a new dataset built on top of Charades by adding sentence temporal annotations.

159 PAPERS • 2 BENCHMARKS

The Distinct Describable Moments (DiDeMo) dataset is one of the largest and most diverse datasets for the temporal localization of events in videos given natural language descriptions. The videos are collected from Flickr and each video is trimmed to a maximum of 30 seconds. The videos in the dataset are divided into 5-second segments to reduce the complexity of annotation. The dataset is split into training, validation and test sets containing 8,395, 1,065 and 1,004 videos respectively. The dataset contains a total of 26,892 moments and one moment could be associated with descriptions from multiple annotators. The descriptions in DiDeMo dataset are detailed and contain camera movement, temporal transition indicators, and activities. Moreover, the descriptions in DiDeMo are verified so that each description refers to a single moment.

159 PAPERS • 3 BENCHMARKS

Youtube-VOS is a Video Object Segmentation dataset that contains 4,453 videos - 3,471 for training, 474 for validation, and 508 for testing. The training and validation videos have pixel-level ground truth annotations for every 5th frame (6 fps). It also contains Instance Segmentation annotations. It has more than 7,800 unique objects, 190k high-quality manual annotations and more than 340 minutes in duration.

156 PAPERS • 10 BENCHMARKS