If you have additions or changes, send an e-mail (remove the "nospam").
This material is presented to ensure timely dissemination of scholarly and technical work.
Copyright and all rights therein are retained by authors or by other copyright holders.
All persons copying this information are expected to adhere to the terms and constraints invoked by each authors copyright.
LabelMe is a web-based image annotation tool that allows researchers to label images and share the annotations with the rest of the community. If you use the database, we only ask that you contribute to it, from time to time, by using the labeling tool.
1521 images with human faces, recorded under natural conditions, i.e. varying illumination and complex background. The eye positions have been set manually.
15,560 pedestrian and non-pedestrian samples (image cut-outs) and 6744 additional full images not containing pedestrians for bootstrapping. The test set contains more than 21,790 images with 56,492 pedestrian labels (fully visible or partially occluded), captured from a vehicle in urban traffic.
The dataset FlickrLogos-32 contains photos depicting logos and is meant for the evaluation of multi-class logo detection/recognition as well as logo retrieval methods on real-world images. It consists of 8240 images downloaded from Flickr.
30000+ frames with vehicle rear annotation and classification (car and trucks) on motorway/highway sequences.
Annotation semi-automatically generated using laser-scanner data.
Distance estimation and consistent target ID over time available.
Phos is a color image database of 15 scenes captured under different illumination conditions.
More particularly, every scene of the database contains 15 different images: 9 images captured under various strengths of uniform illumination, and 6 images under different degrees of non-uniform illumination.
The images contain objects of different shape, color and texture and can be used for illumination invariant feature detection and selection.
California-ND contains 701 photos taken directly from a real user's personal photo collection, including many challenging non-identical near-duplicate cases, without the use of artificial image transformations. The dataset is annotated by 10 different subjects, including the photographer, regarding near duplicates.
This dataset contains 12,995 face images collected from the Internet. The images are annotated with (1) five facial landmarks, (2) attributes of gender, smiling, wearing glasses, and head pose.
WIDER FACE dataset is a face detection benchmark dataset with images selected from the publicly available WIDER dataset. It contains 32,203 images and 393,703 face annotations.
Multiple sequences recorded in two different indoor rooms, using both omnidirectional and perspective cameras, containing people in a variety of situations (people walking, standing, and sitting). Both annotated and non-annotated sequences are provided, where ground truth is point-based. In total, more than 100,000 annotated frames are available.
It contains more than 40.000 images and 230 000 annotated traffic lights and is the largest database for traffic light detection so far containing bounding box labels, track identities and furthermore the following attributes: phase, pictogram, relevancy, occlusion, number of light units and orientation.
A dataset for testing object class detection algorithms. It contains 255 test images and features five diverse shape-based classes (apple logos, bottles, giraffes, mugs, and swans).
A dataset for Attribute Based Classification. It consists of 30475 images of 50 animals classes with six pre-extracted feature representations for each image.
The USAA dataset includes 8 different semantic class videos which are home videos of social occassions which feature activities of group of people. It contains around 100 videos for training and testing respectively. Each video is labeled by 69 attributes. The 69 attributes can be broken down into five broad classes: actions, objects, scenes, sounds, and camera movement.
This database contains 18000 video frames of 640x480 resolution from 60 video sequences, each of which recorded from a different subject (31 female and 29 male).
Video data sets to train machines to recognise objects in our environment. e-VDS35 has 35 classes and a total of 2050 videos of roughly 10 seconds each.
The PubFig database is a large, real-world face dataset consisting of 58,797 images of 200 people collected from the internet. Unlike most other existing face datasets, these images are taken in completely uncontrolled situations with non-cooperative subjects.
The data set contains 3,425 videos of 1,595 different people. The shortest clip duration is 48 frames, the longest clip is 6,070 frames, and the average length of a video clip is 181.3 frames.
The Microsoft Research Cambridge-12 Kinect gesture data set consists of sequences of human movements, represented as body-part locations, and the associated gesture to be recognized by the system.
This dataset contains 250 pedestrian image pairs + 775 additional images captured in a busy underground station for the research on person re-identification.
Face tracks, features and shot boundaries from our latest CVPR 2013 paper. It is obtained from 6 episodes of Buffy the Vampire Slayer and 6 episodes of Big Bang Theory.
ChokePoint is a video dataset designed for experiments in person identification/verification under real-world surveillance conditions. The dataset consists of 25 subjects (19 male and 6 female) in portal 1 and 29 subjects (23 male and 6 female) in portal 2.
Over 110,000 photographic reproductions of the artworks exhibited in the Rijksmuseum (Amsterdam, the Netherlands). Offers four automatic visual recognition challenges consisting of predicting the artist, type, material and creation year. Includes a set of baseline features, and offer a baseline based on state-of-the-art image features encoded with the Fisher vector.
Treadmill gait datasets composed of 34 subjects with 9 speed variations, 68 subjects with 68 subjects, and 185 subjects with various degrees of gait fluctuations.
The FaceScrub dataset is a real-world face dataset comprising 107,818 face images of 530 male and female celebrities detected in images retrieved from the Internet.
The images are taken under real-world situations (uncontrolled conditions). Name and gender annotations of the faces are included.
ECCV 2018: Open Museum Identification Challenge dataset, photos of exhibits captured in 10 distinct exhibition spaces of several museums which showcase paintings, timepieces, sculptures, glassware, relics, science exhibits, natural history pieces, ceramics, pottery, tools and indigenous crafts.
The set was recorded in Zurich, using a pair of cameras mounted on a mobile platform. It contains 12'298 annotated pedestrians in roughly 2'000 frames.
MIT traffic data set is for research on activity analysis and crowded scenes. It includes a traffic video sequence of 90 minutes long. It is recorded by a stationary camera.
This dataset consists of more than 22,000 images of 24 people which are captured by 16 cameras installed in a shopping mall "Shinpuh-kan". All images are manually cropped and resized to 48x128 pixels, grouped into tracklets and added annotation.
Low-resolution RGB videos + ground truth trajectories from multiple fixed and moving cameras monitoring the same scenes (indoor and outdoor) to improve object tracking and matching.
This dataset contains videos of crowds and other high density moving objects. The videos are collected mainly from the BBC Motion Gallery and Getty Images website. The videos are shared only for the research purposes. Please consult the terms and conditions of use of these videos from the respective websites.
Contains hand-labelled pixel annotations for 38 groups of images, each group containing a common foreground. Approximately 17 images per group, 643 images total.
Image segmentation and boundary detection. Grayscale and color segmentations for 300 images, the images are divided into a training set of 200 images, and a test set of 100 images.
The dataset consists of video sequences recorded in urban traffic. The dataset consists of 5000 rectified stereo image pairs. 500 frames come with pixel-level semantic class annotations into 5 classes: ground, building, vehicle, pedestrian, sky. Dense disparity maps are provided as a reference.
LASIESTA is composed by many real indoor and outdoor sequences organized in diferent categories, each of one covering a specific challenge in moving object detection strategies.
For the CAVIAR project a number of video clips were recorded acting out the different scenarios of interest. These include people walking alone, meeting with others, window shopping, entering and exitting shops, fighting and passing out and last, but not least, leaving a package in a public place.
Multiple sequences recorded in two different indoor rooms, using both omnidirectional and perspective cameras, containing people in a variety of situations (people walking, standing, and sitting). Both annotated and non-annotated sequences are provided, where ground truth is point-based. In total, more than 100,000 annotated frames are available.
UNICITY consists of 58k images collected from 65 recorded sequences with one or two people performing different behaviors including attacks and trickeries. It also provides full annotation of people such as the location of head and shoulders.
24 scenarios recorded with 8 IP video cameras. The first 22 first scenarios contain a fall and confounding events, the last 2 ones contain only confounding events.
This dataset consists of 8000+ images of professional footballers during a match of the Allsvenskan league. It consists of two parts: one with ground truth pose in 2D and one with ground truth pose in both 2D and 3D.
This dataset includes: camera calibration information, raw input images we have captured, radially undistorted, rectified, and cropped images, depth maps resulting from our reconstruction and propagation algorithm, depth maps computed at each available view by the reconstruction algorithm without the propagation applied.
24 synthetic scenes. Available data per scene: 9x9 input images (512x512x3) , ground truth (disparity and depth), camera parameters, disparity ranges, evaluation masks.
This dataset consists of a set of actions collected from various sports which are typically featured on broadcast television channels such as the BBC and ESPN. The video sequences were obtained from a wide range of stock footage websites including BBC Motion gallery, and GettyImages.
This dataset features video sequences that were obtained using a R/C-controlled blimp equipped with an HD camera mounted on a gimbal.The collection represents a diverse pool of actions featured at different heights and aerial viewpoints. Multiple instances of each action were recorded at different flying altitudes which ranged from 400-450 feet and were performed by different actors.
The dataset was captured by a Kinect device. There are 12 dynamic American Sign Language (ASL) gestures, and 10 people. Each person performs each gesture 2-3 times.
Contains six types of human actions (walking, jogging, running, boxing, hand waving and hand clapping) performed several times by 25 subjects in four different scenarios: outdoors, outdoors with scale variation, outdoors with different clothes and indoors.
Hollywood-2 datset contains 12 classes of human actions and 10 classes of scenes distributed over 3669 video clips and approximately 20.1 hours of video in total.
This dataset contains 5 different collective activities : crossing, walking, waiting, talking, and queueing and 44 short video sequences some of which were recorded by consumer hand-held digital camera with varying view point.
The dataset is designed to be realistic, natural and challenging for video surveillance domains in terms of its resolution, background clutter, diversity in scenes, and human activity/event categories than existing action recognition datasets.
Collected from various sources, mostly from movies, and a small proportion from public databases, YouTube and Google videos. The dataset contains 6849 clips divided into 51 action categories, each containing a minimum of 101 clips.
Fully annotated dataset of RGB-D video data and data from accelerometers attached to kitchen objects capturing 25 people preparing two mixed salads each (4.5h of annotated data). Annotated activities correspond to steps in the recipe and include phase (pre-/ core-/ post) and the ingredient acted upon.
A Kinect dataset for hand detection in naturalistic driving settings as well as a challenging 19 dynamic hand gesture recognition dataset for human machine interfaces.
This dataset consists of seven meal-preparation activities, each performed by 10 subjects. Subjects perform the activities based on the given cooking recipes.
The dataset consists of four temporally synchronized data modalities. These modalities include RGB videos, depth videos, skeleton positions, and inertial signals (3-axis acceleration and 3-axis angular velocity) from a Kinect RGB-D camera and a wearable inertial sensor for a comprehensive set of 27 human actions.
"NTU RGB+D" contains 60 action classes and 56,880 video samples. "NTU RGB+D 120" extends "NTU RGB+D" by adding another 60 classes and another 57,600 video samples, i.e., "NTU RGB+D 120" has 120 classes and 114,480 samples in total.
These two datasets both contain RGB videos, depth map sequences, 3D skeletal data, and infrared (IR) videos for each sample. Each dataset is captured by three Kinect V2 cameras concurrently.
Contains 91,793 faces manually labeled with expressions. Each of the face images was manually annotated as one of the seven basic expression categories: “angryâ€, “disgustâ€, “fearâ€, “happyâ€, “sadâ€, “surpriseâ€, or “neutralâ€.
This dataset includes 214971 annotated depth images of hands captured by a RealSense RGBD sensor of hand poses. Annotations: per pixel classes, 6D fingertip pose, heatmap. Images -> Train: 202198, Test: 10000, Validation: 2773. Recorded at GIP Lab, Technion.
More than 40.000 scans of people very accurately registered. Scans contain texture so synthetic videos/images are easy to generate. See also Dyna: A Model of Dynamic Human Shape in Motio.
About 10.000 scans of people in clothing and the estimated body shape of people underneath. Scans contain texture so synthetic videos/images are easy to generate.
Chictopia dataset with additional processed annotations (face) and SMPL body model fits to the images. The copyright of the images belongs to the original authors of Chictopia.
DB Contains 100 examples with the uncompressed frames, up to the 10th frame after the appearance of the 8th cell; a text file with the trajectories of all the cells, from appearance to division; a movie file showing the trajectories of the cells.
Miracl-VC1 is a lip-reading dataset including both depth and color images.
Fifteen speakers positioned in the frustum of a MS Kinect sensor and utter ten times a set of ten words and ten phrases.
Home
Changelog
Forum
RSS
Twitter