Some datasets and evaluation tools are provided on this page for four different computer vision and computer graphics problems. Population counting Line-network extraction 3D-object detection Surface reconstruction Surface reconstruction We focus here on the well-known surface reconstruction problem from point clouds. The datasets include simple synthetic objects, but also real-world urban scenes containing various types of geometric structures. The file format for both point clouds and triangular meshes is ply. The software Meshlab can be used to read and convert these files. Evaluation of the results: The results are evaluated according to the error to the ground truth and/or the input point cloud (Hausdorff distance), and time. Meshlab can be used to compute the Hausdorff distance between your results and the given input point clouds / reference meshes. Smooth_cube (100K points) Distorted_cube (50K points) Maya_temple (2M points) Maya_temple_cor (0.83M points) Empire_state_building_cor (1.2M points) Lans_le_Villard (1.23M points) Cathedral_facade (1.45M points) 3D-object detection The goal consists in detecting a specific class of 3D-objects from point clouds. The file format for point clouds is ply. The software Meshlab can be used to read and convert the files. Evaluation of the results: The results can be evaluated according to criteria: the estimated number of objects, and the precision of their position. You can use this Matlab code to evaluate your results from the following datasets. Urban_small (140K points) Urban (2.27M points)
The ICG Multi-Camera and Virtual PTZ dataset contains the video streams and calibrations of several static Axis P1347 cameras and one panoramic video from…
calibration, camera, crowd, detection, graz, multitarget, multiview, network, object, outdoor, panorama, pedestrian, tracking, videoThe 1DSfM Landmarks is a collection of community-based image reconstruction by Kyle Wilson and is comprised of 14 datasets with comparison to bundler grou…
3d, benchmark, city, groundtruth, landmark, reconstruction, urbanThe Daimler Mono Pedestrian Detection Benchmark dataset contains a large training and test set. The training set contains 15.560 pedestrian samples (image…
detection, mono, object, outdoor, pedestrian, scale, urbanThe Mall dataset was collected from a publicly accessible webcam for crowd counting and profiling research. Ground truth: Over 60,000 pedestrians were …
counting, crowd, detection, indoor, pedestrian, tracking, video, webcamThe QMUL Junction dataset is a busy traffic scenario for research on activity analysis and behavior understanding. Video length: 1 hour (90000 frames)…
behavior, counting, crowd, detection, motion, pedestrian, tracking, videoThis ETHZ CVL RueMonge 2014 dataset used for 3D reconstruction and semantic mesh labelling for urban scene understanding. It was first published in [1] …
3d, architecture, benchmark, classification, code, mesh, outdoor, paris, pointcloud, recognition, reconstruction, segmentation, semantic, source, urbanThe High Definition Analytics (HDA) dataset is a multi-camera High-Resolution image sequence dataset for research on High-Definition surveillance: Pedestr…
benchmark, camera, detection, high-definition, human, indoor, lisbon, multiview, network, pedestrian, re-identification, surveillance, tracking, videoThe Freiburg-Berkeley Motion Segmentation Dataset (FBMS-59) is an extension of the BMS dataset with 33 additional video sequences. A total of 720 frames i…
benchmark, groundtruth, motion, object, pedestrian, segmentation, tracking, videoLASIESTA is composed by many real indoor and outdoor sequences organized in different categories, each of one covering a specific challenge in moving obje…
background, camera, challenge, dataset, detection, foreground, groundtruth, motion, object, stationary, subtractionThe PASCAL VOC is augmented with segmentation annotation for semantic parts of objects. For example, for the person category, we provide segmentation mask…
detection, human, object, part, pascal, pedestrian, recognition, segmentation, semanticWe present a new large-scale dataset that contains a diverse set of stereo video sequences recorded in street scenes from 50 different cities, with high q…
car, cities, detection, pedestrian, person, segmentation, semantic, stereo, urban, video, weaklyThe MSR Action datasets is a collection of various 3D datasets for action recognition. See details http://research.microsoft.com/en-us/um/people/zliu/a…
3d, action, detection, recognition, reconstruction, videoThe ICG Lab 6 (Multi-Camera Multi-Object Tracking) dataset contains 6 indoor people tracking scenarios recorded at our laboratory using 4 static Axis P134…
calibration, camera, detection, evaluation, graz, laboratory, multiview, object, pedestrian, segmentation, trackingThis dataset package contains the software and data used for Detection-based Object Labeling on the RGB-D Scenes Dataset as implemented in the paper: De…
3d, depth, indoor, kinect, object, recognition, reconstructionA 66 stereo pairs dataset with their subpixel ground truths. The construction and improvement of algorithms for subpixel stereovision requires very prec…
3d, depth, groundtruth, noise, pointcloud, stereo, stereovision, subpixelISPRS and EuroSDR - Benchmark on High Density Aerial Image Matching Background and Scope of the project Innovations in matching algorithms as well as …
3d, aerial, benchmark, city, germany, multiview, photogrammetry, reconstruction, switzerland, urbanThe Video Segmentation Benchmark (VSB100) provides ground truth annotations for the Berkeley Video Dataset, which consists of 100 HD quality videos divide…
benchmark, groundtruth, motion, object, pedestrian, segmentation, tracking, videoThe ICG Multi-Camera datasets consist of Easy Data Set (just one person) Medium Data Set (3-5 persons, used for the experiments) Hard Data Set (crowd…
calibration, camera, detection, graz, indoor, multitarget, multiview, object, pedestrian, tracking, videoThe crowd datasets are collected from a variety of sources, such as UCF and data-driven crowd datasets. The sequences are diverse, representing dense crow…
anomaly, crowd, detection, human, pedestrian, scene, understanding, videoThe city planar and non-planar datset consists of urban scenes accompanied by text files describing the plane/non-plane locations. Training Set (Univer…
3d, building, detection, estimation, plane, urbanThe New College Data Set contains 30GB of data intended for use by the mobile robotics and vision research communities. Our anticipated users are parties …
3d, navigation, odometry, panorama, path, reconstruction, stereo, urbanThe UrbanStreet dataset used in the paper can be downloaded here [188M] . It contains 18 stereo sequences of pedestrians taken from a stereo rig mounted o…
detection, human, multitarget, pedestrian, recognition, segmentation, tracking, urban, videoThe Daimler Mono Pedestrian Classification Benchmark dataset consists of two parts: a base data set. The base data set contains a total of 4000 pedestri…
classification, illumination, object, outdoor, pedestrian, scale, urbanThe Robotic 3D Scan Repository from Osnabrueck contains 23 different datasets showing a veriaty of 3D scans for objects, humans, cities, university campus…
3d, aerial, bremen, city, germany, heat, human, laser, lidar, osnabrueck, reconstruction, scan, urbanFor the first few decades of the fields existence, computer vision has been focused on algorithmic, logical approaches to perception. But it was only with…
3d, depth, indoor, kinect, object, recognition, reconstructionThe Farman Institute 3D Point Sets dataset contains 11 objects by a 3D laser scanner. This dataset was peer-reviewed by Image Processing On Line: Farman I…
3d, laser, model, object, point, reconstruction, scannerThe KTH Multiview Football dataset contains 771 images of football players includes images taken from 3 views at 257 time instances 14 annotated body join…
camera, detection, game, multitarget, multiview, object, outdoor, pedestrian, pose, recognition, soccer, trackingThe Landmark 1000 or 1k dataset is a collection of the top 1000 popular flickr landmarks mined from flickr. It is maintained by Noah Snavely and publish…
3d, estimation, landmark, location, pointcloud, pose, reconstruction, worldThe Ford Car dataset is joint effort of Pandey et al. (for collecting images, Lidar points, calibration etc.) and us (for annotation of 2D and 3D objects)…
3d, car, detection, groundtruth, lidar, sfmThe Symmetry Facades dataset contains 9 building facades with multiple images. It used for coupled symmetry and structure from motion detection. Couple…
3d, building, facade, reconstruction, repetition, sfm, symmetry, urbanThe Eurasian Cities dataset contains 103 images of outdoor urban scenes taken in Eurasian cities. It is annotated with horizontal and vertical vanishing p…
geometry, line, manhattan, outdoor, point, pose, reconstruction, urban, vanishingThe Leuven Stereo Scene dataset is a scene and depth dataset. There exist two variants of this dataset - a CVPR 2007 paper [1] by Leibe et al. for detecti…
3d, depth, leuven, reconstruction, segmentation, semantic, sfm, stereo, urbanThe Yotta dataset consists of 70 images for semantic labeling given in 11 classes. It also contains multiple videos and camera matrices for 14km or drivin…
3d, camera, classification, reconstruction, segmentation, semantic, urban, videoThe CVC Partial Occlusion Virtual Pedestrian datasets (CVC-01 to CVC-06) cover a range of scenarios of occluded pedestrians generated in a virtual and rea…
classification, detection, occlusion, pedestrian, synthetic, tracking, urbanISPRS / EuroSDR Benchmark for Multi-Platform Photogrammetry In these pages you can get information about the BENCHMARK FOR MULTI-PLATFORM PHOTOGRAMMETRY…
3d, aerial, benchmark, city, germany, multiview, photogrammetry, reconstruction, switzerland, urbanYahoo Flickr Creative Commons 100M (YFCC100M) dataset contains a list of photos and videos. This list is compiled from data available on Yahoo! Flickr. Al…
3d, clustering, community, detection, flickr, image, internet, landmark, recognition, reconstruction, socialThe Traffic Video dataset consists of X video of an overhead camera showing a street crossing with multiple traffic scenarios. The dataset can be downlo…
detection, overhead, road, tracking, traffic, urban, video, viewThe TRaffic ANd COngestionS (TRANCOS) dataset, a novel benchmark for (extremely overlapping) vehicle counting in traffic congestion situations. It consist…
car, detection, highway, object, spain, traffic, transportation, urban, vehicleThe Where Who Why (WWW) dataset provides 10,000 videos with over 8 million frames from 8,257 diverse scenes, therefore offering a superior comprehensive d…
crowd, detection, flow, optical, pedestrian, recognition, surveillance, videoThe Caltech Lanes dataset includes four clips taken around streets in Pasadena, CA at different times of day. The archive below includes 1225 individual…
caltech, detection, lane, pasadena, road, urbanThe TUD Crossing dataset from Micha Andriluka, Stefan Roth and Bernt Schiele consists of 201 images with 1008 highly overlapping pedestrians with signific…
detection, multitarget, overlap, pedestrian, segmentation, sideview, tracking, urbanISPRS Test Project on Urban Classification and 3D Building Reconstruction The ISPRS working group III/4 announces the release of the 2D semantic labelin…
3d, building, city, classification, recognition, reconstruction, semantic, urbanThis UIUC Cars dataset by Shivani Agarwal, Aatif Awan and Dan Roth contains images of side views of cars for use in evaluating object detection algorithms…
car, detection, recognition, scale, sideview, urbanThe UCF Person and Car VideoSeg dataset consists of six videos with groundtruth for video object segmentation. Surfing, jumping, skiing, sliding, big ca…
camera, groundtruth, model, motion, object, segmentation, videoThe Semantic Description of Human Activities (SDHA) was a contest at ICPR 2010. The contest is composed of three different types of activity recognitio…
aerial, crowd, object detection, object tracking, occlusion, overlap, pedestrian, trajectoryDaimler Stereo Pedestrian Detection Benchmark C. Keller, M. Enzweiler, and D. M. Gavrila, A New Benchmark for Stereo-based Pedestrian Detection, Proc. …
object detection, pedestrian, urbanPenn-Fudan Pedestrian Detection and Segmentation
background, detection, motion, pedestrian, segmentationThis site is dedicated to provide datasets for the Robotics community with the aim to facilitate result evaluations and comparisons. The datasets presente…
3d, city, laser, nature, urbant is composed of food intake movements, recorded with Kinect V1 (320240 depth frame resolution), simulated by 35 volunteers for a total of 48 tests. The d…
age, behavior, food, groundtruth, human, intake, kinect, monitoring, pointcloud, trackingThe mirror symmetry database contains 176 single-symmetry and 63 multyple-symmetry images (.png files) with accompanying ground-truth annotations (.mat fi…
detection, groundtruth, mirror, symmetryThe MOT Challenge is a framework for the fair evaluation of multiple people tracking algorithms. In this framework we provide: - A large collection of d…
3d, benchmark, benhttp://motchallenge.net/chmark, dataset, evaluation, multiple, pedestrian, people, surveillance, target, tracking, videoThe Swedish Traffic Sign Recognition provides Matlab code for parsing the annotation files and displaying the results. Part0 for each set contains the ann…
city, detection, recognition, sign, traffic, urbanThe 2D-3D-S dataset provides a variety of mutually registered modalities from 2D, 2.5D and 3D domains, with instance-level semantic and geometric annotati…
3d, building, depth, indoor, large-scale, normal, panorama, reconstruction, segmentation, semanticVoxel Based Dataset for Systematic 3D reconstruction by artificial neural networks (ANNs). A synthetic scalable cube dataset for training, testing and v…
3d, deep learning, reconstruction, sfm, synthetic city urbanThe BIWI Walking Pedestrians (EWAP) dataset shows walking pedestrians in busy scenarios from a bird eye view. Manually annotated. Data used for training i…
aerial, crowd, object detection, object tracking, occlusion, overlap, pedestrian, trajectoryThe CALTECH 256 dataset by Li Fei-Fei contains 30607 images for 256 categories.
centered, classification, detection, image, object, sceneThe Aspect Layout dataset is designed to allow evaluation of object detection for aspect ratios in perspective images. Author text: In this project we…
aspect, detection, layout, object, perspective, ratioThe Longterm Pedestrian dataset consists of images from a stationary camera running 24 hours for 7 days at about 1 fps. It used for adaptive detection an…
background, change, coffee, detection, graz, illumination, indoor, multitarget, pedestrian, robustThe YouTube-Objects dataset is composed of videos collected from YouTube by querying for the names of 10 object classes. It contains between 9 and 24 vide…
detection, flow, object, optical, segmentation, videoThe Microsoft COCO (mscoco) is an image recognition and segmentation dataset which contains more 300k images for more than 70 categories. Other features…
benchmark, context, detection, object, recognition, segmentation, semanticMany different labeled video datasets have been collected over the past few years, but it is hard to compare them at a glance. So we have created a handy …
action, benchmark, classification, detection, object, recognition, videoThe PETS 2009 dataset contains 3 parts showing multi-view sequences containing pedestrians walking in an outdoor environment. The parts are used for perso…
detection, frontview, human, occlusion multitarget, outdoor, overlap, pedestrian, trackingAbstract Scene understanding has (again) become a focus of computer vision research, leveraging advances in detection, context modeling, and tracking. In…
3d, car, classification, pedestrian, scene, segmentation, semantic, understandingThe 3DVis dataset includes a set of 12 heterogeneous scenes for testing 3D scene registration and analysis methods. Models include homogeneous shapes, rep…
3d, matching, reconstruction, registration, shape, symmetrySVHN is a real-world image dataset for developing machine learning and object recognition algorithms with minimal requirement on data preprocessing and fo…
classification, detection, number, real, recognition, streetside, streetview, text, urban, worldThe Inria Aerial Image Labeling addresses a core topic in remote sensing: the automatic pixelwise labeling of aerial imagery (link to paper). Dataset fe…
aerial, building, city, footprint, groundtruth, house, segmentation, semantic, urbanThe Lane Level Localization dataset was collected on a highway in San Francisco with the following properties: * Reasonable traffic * Multiple lane hig…
3d, autonomous, benchmark, car, driving, gps, localization, map, road, videoA dataset acquired with 3 synchronized sensors (Primesense Carmine 1.09, Microsoft Kinect v2, Canon IXUS 950 IS), featuring: * 30 industry-relevant obje…
3d, estimation, object, pose, rgbd, texture-lessThe York Urban Line Segment Database is a compilation of 102 images (45 indoor, 57 outdoor) of urban environments consisting mostly of scenes from the cam…
geometry, manhattan, outdoor, pose estimation, reconstruction, urban, vanishing pointThe Caltech Pedestrian Dataset consists of approximately 10 hours of 640x480 30Hz video taken from a vehicle driving through regular traffic in an urban e…
object detection, pedestrian, urbanWe collected a video dataset, termed ChokePoint, designed for experiments in person identification/verification under real-world surveillance conditions u…
clustering, detection, face, human, identification, multiview, pedestrian, real, recognition, sequence, surveillance, worldThe HCI 4D Lightfields dataset contains 11 objects with corresponding lightfields for depth estimation. Datasets can be downloaded individually below. F…
3d, 4d, benchmark, depth, evaluation, lightfield, reconstructionThe BEOID dataset includes object interactions ranging from preparing a coffee to operating a weight lifting machine and opening a door. The dataset is re…
3d, egocentric, interaction, object, pose, tracking, videoThe RGB-D Person Re-identification dataset is for person re-identification using depth information. The main motivation is that the standard techniques (s…
3d, classification, depth, identification, pedestrian, shapeAn Annotated Dataset For Near-Duplicate Detection In Personal Photo Collections Managing photo collections involves a variety of image quality assessmen…
copyright, detection, duplicate, groundtruth, retrievalBelgiumTS is a large dataset with 10000+ traffic sign annotations, thousands of physically distinct traffic signs. 4 video sequences recorded with 8 high …
belgium, calibration, camera, classification, road, sign, traffic, urbanA new large-scale PEdesTrian Attribute (PETA) dataset. The dataset is by far the largest of its kind, covering more than 60 attributes on 19000 images. I…
crowd, pedestrianThe 3D shape description dataset consists of multiple sub-datasets Descriptor Matching - Dataset 1 & 2 (Stanford) These datasets, created from some of …
3d, benchmark, description, matching, reconstruction, registration, shapeAt Udacity, we believe in democratizing education. How can we provide opportunity to everyone on the planet? We also believe in teaching really amazing an…
autonomous, car, classification, detection, driving, recognition, robot, segmentation, street, synthetic, time, urban, videoThis dataset is for people tracking in wide baseline camera networks and was designed as a contest at ICPR 2012. The contest consists of two challenges…
aerial, crowd, object detection, object tracking, occlusion, overlap, pedestrian, trajectoryParis-rue-Madame dataset contains 3D Mobile Laser Scanning (MLS) data from rue Madame, a street in the 6th Parisian district (France). The test zone conta…
3d, classification, laser, pointcloud, segmentation, semanticMIT Pedestrian dataset from Papageorgiou and Poggio [IJCV2000] contains 509 training and 200 test images of pedestrians in city scenes (plus left-right re…
boundingbox, frontview, object detection, pedestrian, people, urbanThe Notre Dame de Paris dataset used for 3D SfM reconstruction and contains 715 images provided by Noah Snavely. There are also version for NotreDame b…
3d, 3d reconstruction, flickr, frontview, landmark, limited, paris, pointcloud, sfmSceneNet RGB-D is dataset comprised of 5 million Photorealistic Images of Synthetic Indoor Trajectories with Ground Truth. It expands the previous work of…
3d, indoor, lighting, navigation, reconstruction, rendering, robot, scene, segmentation, slam, synthetic, trajectoryThe PSU HUB dataset is a detection, tracking dataset. Ground truth trajectory and grouping information for pedestrians walking in the PSU student union bu…
crowd, object detection, object tracking, occlusion, overlap, pedestrian, trajectoryThe TU Berlin Multi-Object and Multi-Camera Tracking Dataset (MOCAT) is a synthetic dataset to train and test tracking and detection systems in a virtual …
animal, detection, evaluation, multi-class, multi-view, pedestrian, synthetic, tracking, vehicleThe German Traffic Sign Recognition Benchmark is a dataset for multi-class detection problem in natural images and do cordially invite you to participate.…
detection, recognition, traffic, traffic sign, urbanThe Oxford RobotCar Dataset contains over 100 repetitions of a consistent route through Oxford, UK, captured over a period of over a year. The dataset cap…
autonomous, car, classification, detection, driving, recognition, robot, segmentation, street, time, urban, video, yearThe SegTrack dataset consists of six videos (five are used) with ground truth pixelwise segmentation (6th penguin is not usable). The dataset is used for …
camera, flow, groundtruth, model, motion, object, optical, proposal, segmentation, stationary, videoDaimler Multi-Cue, Occluded Pedestrian Classification Benchmark Training and test samples have a resolution of 48 x 96 pixels with a 12-pixel border aro…
image classification, object detection, pedestrian, urbanThe SUNCG dataset is a Large 3D Model Repository for Indoor Scenes. SUNCG is an ongoing effort to establish a richly-annotated, large-scale dataset of…
3d, indoor, layout, object, realism, recognition, rendering, room, scene, segmentation, syntheticISPRS Test Project on Urban Classification, 3D Building Reconstruction and Semantic Labeling. In this part of our working group site you will get further …
3d, aerial, benchmark, canada, city, germany, multiview, photogrammetry, recognition, segmentation, semantic, urbanThe dataset contains 15 documentary films that are downloaded from YouTube, whose durations vary from 9 minutes to as long as 50 minutes, and the total nu…
detection, object, videoThis dataset provides a collection of web images and 3D models for research on landmark recognition (especially for methods based on 3D models). We hope i…
3d, classification, codebook, feature, flickr, landmark, matching, recognition, reconstruction, retrievalThe Comprehensive Cars (CompCars) dataset contains data from two scenarios, including images from web-nature and surveillance-nature. The web-nature data …
attribute, car, classification, fine-grained, object, recognition, urban, vehicleThe Quad 6K dataset is a Structure-from-Motion dataset taken at Arts Quad at Cornell University campus and consists of 6514 images with ground truth posit…
3d gps, 3d reconstruction, groundtruth, landmark, sfm, urbanBelgiumTSC dataset is built for traffic sign classification purposes. Is is a subset of BelgiumTS dataset and contains cropped images around annotations f…
belgium, classification, road, sign, traffic, urbanScanNet is an RGB-D video dataset containing 2.5 million views in more than 1500 scans, annotated with 3D camera poses, surface reconstructions, and insta…
3d, cad, indoor, layout, object, realism, recognition, rendering, room, scene, segmentation, syntheticWe provide the three datasets used for testing our system for our ICCV 2007 publication, including annotations. Data was recorded using a pair of AVT Marl…
crowd, object detection, object tracking, occlusion, overlap, pedestrian, trajectoryThe Rent3D dataset comprises floorplans and images. The goal of this work is to enable a 3D virtual-tour of an apartment given a small set of monocular im…
apartment, building, floorplan, indoor, layout, reconstruction, urbanThe FlickrLogos-32 dataset contains photos showing brand logos and is meant for the evaluation of multi-class logo recognition as well as logo retrieval m…
classification brand boundingbox, detection, flickr, image, logo, machine learning, object recognition, retrievalTh EPFL Multi-View Car dataset contains 20 sequences of cars as they rotate by 360 degrees. There is one image approximately every 3-4 degrees. Using the …
car, detection, estimation, multiview, pose, rotationThe MSR RGB-D Dataset 7-Scenes dataset is a collection of tracked RGB-D camera frames. The dataset may be used for evaluation of methods for different app…
depth, kinect, location, reconstruction, tracking, videoThe Kendall Square webcam dataset consists of two streams for one sunny day and one cloudy day of a city square. It is used for tracking and analyzing col…
appearance, change, color, detection, sky, weather, webcamThe TUD Pedestrians training dataset from Micha Andriluka, Stefan Roth and Bernt Schiele consists of 210 and 400 training images with X pedestrians with s…
object detection, pedestrian, segmentation, sideviewUnlike the previous SHREC contests, the objective of this SHREC 2012 contest is to evaluate the performance of 3D-mesh segmentation techniques instead of …
3d, mesh, part, segmentationThe PETS 2016 IPATCH dataset contains a set of fourteen multi camera recordings (visible, themal) collected off the coast of Brest, France, in collaborati…
boat, detection, gps, maritime, multimodal, radar, thermal, tracking, vessel, visibleThe PASCAL VOC Challenge datasets by Mark Everingham is a yearly dataset which has a central evaluation server and the final test data is not released. Th…
airplane, animal, building, car, chair, object detection, object pose, object segmentation, pedestrianThe Street View Text (SVT) dataset contains 647 words and 3796 letters in 249 images harvested from Google Street View. The dataset is more challenging …
classification, outdoor, text detection, text recognition, urbanThe Caltech Buildings dataset consists of images taken for 50 buildings around the Caltech campus. Five different images were taken for each building from…
building, caltech, hierarchy, retrieval, taxonomy, urbanChairGest is an open challenge / benchmark. The task consists in spotting and recognizing gestures from multiple synchronized sensors: 1 Kinect and 4 Xse…
benchmark, detection, gesture, human, kinect, recognitionThe Stanford 40 Actions dataset contains images of humans performing 40 actions. In each image, we provide a bounding box of the person who is performing …
action, boundingbox, detection, human, recognitionThe Paris dataset consists of 6412 images. Images have high resolution and are in JPEG format. http://www.robots.ox.ac.uk/~vgg/data/parisbuildings/pari…
image retrieval, landmark, paris, urbanThe Wide (multiple) Baseline Dataset. 31 image pairs, simultaneously combining several nuisance factors: geometry, illumination, IR-visible, etc. WxBS: …
day, description, detection, feature, ir, matching, night, viewpointIdiap/ETHZ Faces and Poses Dataset dataset by L. Jie, B. Caputo and V. Ferrari contains 1703 image-caption pairs. [author] Captions contain the names of s…
face detection, object pose, pedestrian, textThe TUD Campus dataset from Micha Andriluka, Stefan Roth and Bernt Schiele consists of 71 images and 303 highly overlapping pedestrians with large scale c…
object detection, object tracking, overlap, pedestrian, segmentation, sideviewThe Yale Face dataset from A. Georghiades contains 5760 single light source images of ten subjects, each shown in 9 poses and 64 illumination setups (lead…
face detection, illumination, pedestrian, pose estimationCalifornia-ND contains 701 photos taken directly from a real user's personal photo collection, including many challenging non-identical near-duplicate cas…
detectionThe Leeds Cows dataset by Derek Magee consists of 14 different video sequences showing a total of 18 cows walking from right to left in front of different…
animal, background, cow, detection, segmentation, videoThe VOT2016 pixel-wise annotations dataset contains pixel-wise per-frame annotations for sequences from VOT2016 dataset. The annotation is in a form of BW…
annotation, mask, object, segmentation, tracking, visual10000 images of natural scenes grabbed on Flickr, with 2695 logos instances cut and pasted from the BelgaLogos dataset.
detectionThis is a subset of the dataset introduced in the SIGGRAPH Asia 2009 paper, Webcam Clip Art: Appearance and Illuminant Transfer from Time-lapse Sequences.…
camera, change, illumination, light, nature, static, time, urban, video, webcamThe All I Have Seen (AIHS) dataset is created to study the properties of total visual input in humans, for around two weeks Nebojsa Jojic wore a camera ca…
3d, clustering, indoor, outdoor, scene, similarity, study, summary, user, videoThe TUD Pedestrians dataset from Micha Andriluka, Stefan Roth and Bernt Schiele [AndrilukaCVPR2008] consists of 250 images with 311 fully visible people w…
object detection, pedestrian, segmentation, sideviewThe eTrims dataset is comprised of two datasets, the 4-Class eTRIMS Dataset with 4 annotated object classes and the 8-Class eTRIMS Dataset with 8 annotate…
procedural reconstruction, semantic segmentation, urbanThe HandNet dataset contains depth images of 10 participants hands non-rigidly deforming infront of a RealSense RGB-D camera. This dataset includes 214…
articulation, classification, detection, fingertip, hand, pose, rgbd, segmentation, videoThe ICG Graz240 dataset consists of 240 buildings with 5400 redundant images with a total of 5542 window instances. Window detection itself is difficult d…
graz, object detection, semantic, semantic segmentation, urbanCMU/VMR Urban Image+Laser dataset contains 372 images linked with 3D laser points projections. There are additional images (due to the laser scanner being…
3d reconstruction, laser, semantic segmentation, sfm, urbanWe share our omnidirectional and panoramic image dataset (with annotations) to be used for human and car detection. Please reach through: http://cvrg.iyt…
car, detection, human, omnidirection, panorama, recognitionThe CALTECH 101 dataset by Li Fei-Fei contains images for 101 categories with about 40 to 800 images per category. Most categories have about 50 images at…
centered, image classification, natural-image, object, sceneThe ECP New York dataset contains 10 manually segmented buildings from New York City, USA. Segmentation evaluating using Dice coefficient is calculated fo…
newyork, procedural reconstruction, semantic, semantic segmentation, urbanThe Weather and Illumination Database (WILD) is an extensive database of high quality images of an outdoor urban scene, acquired every hour over all seaso…
camera, change, depth, estimation, illumination, light, newyork, static, time, urban, video, weather, webcamThe LabelMeFacade dataset contains buildings, windows, sky and a limited number of unlabeled regions (maximally 20% covering of the image). This procedure…
facade, recognition, rectified, segmentation, semantic, urbanThe NYU-Depth V2 data set is comprised of video sequences from a variety of indoor scenes as recorded by both the RGB and Depth cameras from the Microsoft…
depth, kinect, label, reconstruction, semantic segmentationThe dataset captures 25 people preparing 2 mixed salads each and contains over 4h of annotated accelerometer and RGB-D video data. Annotated activities co…
action, activity, classification, detection, recognition, tracking, videoCMP Dataset by Ondra Chum contains 5 million images collected from the internet.
image retrieval, large scale, urbanThis repository contains labeled 3-D point cloud laser data collected from a moving platform in a urban environment. Data are provided for research purpos…
3d reconstruction, laser, semantic segmentation, sfm, urbanThe Oxford Buildings dataset by James Philbin and Andrew Zisserman consists of 5062 images collected from Flickr by searching for particular Oxford landma…
image retrieval, landmark, oxford, urbanCollected in a clothing store. Captured with Kinect (640*480, about 30fps)
detection, trackingThis dataset contains 7 challenging volleyball activity classes annotated in 6 videos from professionals in the Austrian Volley League (season 2011/12). A…
action, activity recognition, analysis, detection, sport, video, volleyballThe Deformed Lattice Detection In Real-World Images dataset is used for regular grid detection. The authors have developed a robust and fast lattice detec…
lattice detection, symmetry, texture segmentation, urbanThe object is a plaster reproduction of Temple of the Dioskouroi in Agrigento, Sicily. Click on thumbnail for a full-sized (640x480) image. Resolution of …
3d, 3d reconstruction, benchmark, multiview, sfmWe introduce the Shelf dataset for multiple human pose estimation from multiple views. In addition we annotate the body joints in the Campus dataset from …
3d, capture, estimation, human, motion, multiple, pose, viewThe Google Street View dataset contains 62,058 high quality Google Street View images. The images cover the downtown and neighboring areas of Pittsburgh, …
address, google, gps, localization, manhattan, panorama, pittsburgh, retrieval, sphere, streetview, urbanSheffield Building Image Dataset consists of over 3,000 low-resolution images of forty different buildings typically between 70 and 120 images per buildi…
image classification, image retrieval, sheffield, urbanPhos is a color image database of 15 scenes captured under different illumination conditions. More particularly, every scene of the database contains 15 …
detectionCOCO-Stuff augments the COCO dataset with pixel-level stuff annotations for 10,000 images. These annotations can be used for scene understanding tasks lik…
annotation, benchmark, captioning, coco, groundtruth, segmentation, semantic, stuff, thingsPN Learning - How does TLD work? Tracking estimates the object location as long as the object is visible. During tracking all observed patterns of the o…
bike, face, learning, object tracking, pedestrian, singletargetThe Paris Art Deco Facades dataset consists of 79 / 80 images of rectified facades of the architectural style Art Deco, which has different sizes of windo…
architecture, city, facade, grammar, paris, procedural, recognition, segmentation, semantic, urbanThe Google Street View Pittsburgh Research dataset is a street-level image collection provided by Google for research purposes. The dataset provided he…
3d reconstruction, panorama, pittsburgh, sfm, urbanThe Stable Structure from Motion datasets due to size limitations cannot put the images online. Instead here are the tracked image points and the final re…
3d, 3d reconstruction, church, geometry, landmark, robust, sfm, stabilityThe Stanford Dogs dataset contains images of 120 breeds of dogs from around the world. This dataset has been built using images and annotation from ImageN…
classification, detection, dogs, fine-grained categorizationThe Symmetric Bundle Adjustment dataset contains four sequences of the CAB building, Barcelona, Redmond and Capitole for 3D reconstruction considering sym…
3d reconstruction, bundle adjustment, sfm, symmetry, urbanThe Video Summarization (SumMe) dataset consists of 25 videos, each annotated with at least 15 human summaries (390 in total). The data consists of videos…
action, benchmark, event, groundtruth, human, summary, videoThe Graz02 dataset by Andreas Opelt and Axel Pinz contains four categories of images: bikes, people, cars and a single background class. The annotation ha…
background, bike, car, clutter, graz, object detection, pedestrianThe Cambridge-driving Labeled Video Database (CamVid) dataset from Gabriel Brostow [?] contains ten minutes of video footage and corresponding semanticall…
3d reconstruction, depth, semantic, semantic segmentation, sfm, urbanThe object is a plaster dinosaur (stegosaurus). Click on thumbnail for a full-sized (640x480) image. Resolution of ground truth model: 0.00025m (you may w…
3d, 3d reconstruction, benchmark, multiview, sfm1521 images with human faces, recorded under natural conditions, i.e. varying illumination and complex background. The eye positions have been set manuall…
detection10000 images of natural scenes, with 37 different logos, and 2695 logos instances, annotated with a bounding box.
detectionThe UK Bench dataset from Henrik Stewenius and David Nister contains 10200 images of N=2550 groups with each four images at size 640x480. The images are r…
centered, image retrieval, object, rotationThe Ecole Centrale Paris 2010 (Paris 2010) dataset consists of 30 images of densely annotated building facades in seven classes - wall, window, sky, shop,…
paris, procedural reconstruction, semantic, semantic segmentation, urbanLabelMe 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 …
detectionThe multi-modal/multi-view datasets are created in a cooperation between University of Surrey and Double Negative within the EU FP7 IMPART project. The …
3d, action, color, dynamic, emotion, face, human, indoor, lidar, model, multi-mode, multi-view, outdoor, rgbd, videoWe take advantage of our autonomous driving platform Annieway to develop novel challenging real-world computer vision benchmarks. Our tasks of interest ar…
depth, detection tracking, object detection, object tracking, odometry, optical flow, reconstruction, segmentation, semantic car depth, sfm, stereoDatabase contains 798 images of 114 persons, with 7 images per person and is freely available for research purposes. All images were taken in supervised c…
biometry, face, human, illumination, lighting, pedestrian, person, recognitionThe Colosseum and San Marco are two image datasets for dense multiview stereo reconstructions used for evaluating the visual photo realism. The datasets…
3d reconstruction, aerial, flickr, landmark, photo-realism, sfm, streetside, urbanThe CMP Facade dataset consists of facade images assembled at the Center for Machine Perception, which includes 600 rectified images of facades from vario…
classification, facade, recognition, rectification, segmentation, semantic, similarity, structure, urbanThe FaceScrub dataset comprises a total of 107818 unconstrained face images of 530 celebrities crawled from the Internet, with about 200 images per person…
celebrity, detection, face, human, people, recognitionThe Extreme Zoom Dataset. EZD is a 6 image sets with incleasing zoom factor from general scene view to focusing on single detail. MODS: Fast and Robust …
description, detection, feature, matching, viewpoint, zoom15,560 pedestrian and non-pedestrian samples (image cut-outs) and 6744 additional full images not containing pedestrians for bootstrapping. The test set c…
detectionWe present the 2017 DAVIS Challenge, a public competition specifically designed for the task of video object segmentation. Following the footsteps of othe…
benchmark, code, hd, object, quality, resolution, segmentation, tracking, video segmentationThe Berkeley DeepDrive Video Dataset contains 2x order of magnitude more video training data.
autonomous, deep, driving, endtoend, learning, urbanSince its launch in September 1999, Space Imaging IKONOS earth imaging satellite has provided a reliable stream of image data that has become the standard…
3d reconstruction, aerial, photogrammetry, sfm, urbanThe Symmetry set dataset is a collection of images at different illuminations for the purpose of image matching using local symmetry features. Image Mat…
building, feature, illumination, image, lighting, matching, symmetry, urbanContains drawing pages from US patents with manually labeled figure and part labels.
detectionThe DTU Robot dataset consists of color images of 60 scenes acquired in a controlled setup from 119 different positions and under different lighting. For …
feature description, feature detection, feature matching, illumination, reconstruction, sfmThe Graz01 dataset by Andreas Opelt and Axel Pinz contains four types of images: bikes, people, background with no bikes, background with no people.
background, bike, clutter, graz, object detection, occlusion, pedestrianThe INRIA People dataset from Navneet Dalal and Bill Triggs [DalalCVPR2005] consists of training and testing data. The training contains 1805 images and X…
boundingbox, frontview, human, object detection, pedestrian, sideviewThe Synthetic CAD Models dataset consists of X synthetic CAD models for detection (planar) primitives. Efficient RANSAC for Point-Cloud Shape Detection …
3d object, model fitting, primitive, ransac, reconstruction, syntheticIn HouseCraft, we utilize rental ads to create realistic textured 3D models of building exteriors. In particular, we exploit the address of the property a…
building, city, floorplan, house, localization, registration, segmentation, semantic, streetview, urbanGlobal Symmetry Ground-truth for AVA dataset Release Date: 2016 For detailed information, please refer to: Elawady, Mohamed, Ccile Barat, Christophe …
aesthetic, bilateral, detection, global, mirror, reflection, symmetryZuBuD+, created in February 2017 by Federico Magliani (University of Parma), introduces many query images balancing the class evaluated from the previous …
building, image retrieval, landmark, urbanThe We Are Family Stickmen dataset from Eichner and Ferrari contains X images with X people in group photos for human pose estimation with annotated 2D hu…
body part, object pose, pedestrianThe GaTech VideoContext dataset consists of over 100 groundtruth annotated outdoor videos with over 20000 frames for the task of geometric context evalua…
classification, context, geometry, nature, outdoor, segmentation, semantic, supervised, unsupervised, urban, videoRobust Multi-Person Tracking from Mobile Platforms In all cases, data was recorded using a pair of AVT Marlins F033C mounted on a chariot respectively a…
color, pedestrian, sequence, trackingThe Hopkins 155 Dataset has been created with the goal of providing an extensive benchmark for testing feature based motion segmentation algorithms. It co…
motion segmentation, optical flow, stereo estimation, urbanThe Daimler Urban Segmentation Dataset consists of video sequences recorded in urban traffic. The dataset consists of 5000 rectified stereo image pairs wi…
motion, outdoor, segmentation, semantic, stereo, urbanWIDER FACE dataset is a large-scale face detection benchmark dataset with 32,203 images and 393,703 face annotations, which have high degree of variabilit…
detection, face, occlusion, pose, scaleThe automotive multi-sensor (AMUSE) dataset consists of inertial and other complementary sensor data combined with monocular, omnidirectional, high frame …
api, city, image, inertial, streetside, traffic, urban, videoDataset contains 1000 images of 100 persons, with 10 images per person and is freely available. All images were acquired by cropping ears from images from…
biometry, ear, human, lighting, pedestrian, person, recognitionThe 3D Mask Attack Database (3DMAD) is a biometric (face) spoofing database. It currently contains 76500 frames of 17 persons, recorded using Kinect for b…
3d, biometry, emotion, face, frontview, recognition, segmentationThe KU Leuven Facade dataset is used for architectural styles classification. M. Mathias, A. Martinovic, J. Weissenberg, S. Haegler, L. Van Gool: Automa…
architecture, image classification, procedural reconstruction, urbanThe Zurich Building dataset (ZuBud) from Hao Shao, Tomas Svoboda and Luc Van Gool [?] contains 1005 images with 201 buildings each in five views. There is…
image retrieval, procedural, rectification, urban3 datasets: PTZ Tracking, Thermal-visible registration, Single object tracking
pedestrian, ptz, thermal, trackingThe contour patches dataset is a large dataset of images patch matches used for contour detection. References: C. L. Zitnick and D. Parikh The Role o…
contour, detection, edge, image, lowlevel, match, patch, segmentationSince the publicly available face image datasets are often of small to medium size, rarely exceeding tens of thousands of images, and often without age in…
age, biometry, detection, face, imdb, recognition, wikipediaWelcome to the homepage of the gvvperfcapeva datasets. This site serves as a hub to access a wide range of datasets that have been created for projects of…
action, depth, face, human, mesh, multiview, pose, reconstruction, tracking, videoThe PIROPO database (People in Indoor ROoms with Perspective and Omnidirectional cameras) comprises multiple sequences recorded in two different indoor ro…
detection, fisheye, human, indoor, omnidirectional, people, perspective, room, surveillanceThe NYU-Depth data set is comprised of video sequences from a variety of indoor scenes as recorded by both the RGB and Depth cameras from the Microsoft Ki…
depth, kinect, label, reconstruction, semantic segmentationThe Multi-illuminant Image Sequences dataset contains 16 video sequences (13 with single light source and 3 with two global light sources), recorded with…
balance, chromaticity, color, constancy, dichromatic, illumination, light, nature, object, physics, whiteWe developed imaging hardware consisting of a color camera, a thermal camera and a beam splitter to capture the aligned multispectral (RGB color + Thermal…
pedestrian, rgb, thermalPedestrian Color Naming (PCN) dataset contains 14,213 images, each of which hand-labeled with color label for each pixel.
color naming, pedestrian, segmentationInstance recognition from depth data. Contains various challenges of Pose, Clutter, Occlusion and similar looking objects (Bonde, U., Badrinarayanan, V., …
depth, detection, instance, poseZurich City Hall dataset (also CIPA dataset) nformation: Place: City Hall, Zurich, Switzerland Number of Images: 15, 1280 x 1000 pixels Camera: Fuji …
3d reconstruction, photogrammetry, sfm, urban, zurichScene Background Initialization (SBI) dataset The SBI dataset has been assembled in order to evaluate and compare the results of background initializati…
background, benchmark, change, detection, foreground, initializationThe Stanford Background Dataset is a new dataset introduced in Gould et al. (ICCV 2009) for evaluating methods for geometric and semantic scene understand…
classification, geometry, nature, segmentation, semantic, urbanThe San Francisco Landmark Dataset for Mobile Landmark Recognition is a set of images and query images for localization. We present the San Francisco La…
calibration, city, gps, landmark, localization, mobile, retrieval, sanfrancisco, urbanThis is a dataset of rectified facade images and semantic labels. The goal of the annotation is to study the layout of the facades. It contains 50 imag…
graz, procedural reconstruction, semantic, semantic segmentation, urbanContains 6 object categories similar to object categories in Pascal VOC that are suitable for studying the abnormalities stemming from objects.
detectionThe goal of LabelMe is to provide an online annotation tool to build image databases for computer vision research. You can contribute to the database by v…
object detection, outdoor, semantic, semantic segmentation, software, urbanThe CIFAR-10 dataset consists of 60000 32x32 colour images in 10 classes, with 6000 images per class. There are 50000 training images and 10000 test image…
color, image classification, object, patch, scene, tiny30000+ frames with vehicle rear annotation and classification (car and trucks) on motorway/highway sequences. Annotation semi-automatically generated usin…
detectionThe ECP Paris 2011 dataset consists of 104 images taken from rue Monge in the fifth district of Paris, we kept only 20 for training and 10 for testing. …
paris, procedural reconstruction, semantic, semantic segmentation, urbanClassification/Detection Competitions, Segmentation Competition, Person Layout Taster Competition datasets
classification, detectionThe Pedestrian Parsing dataset contains 3,673 images from 171 videos of different Surveillance Scenes (PPSS), where 2,064 images are occluded and 1,609 ar…
parsing, pedestrian, segmentation15 wide baseline stereo image pairs with large viewpoint change, provided ground truth homographies. Image size (~1000x700 pixels, RGB) D. Mishkin and…
description, detection, feature, matching, viewpoint, wide baseline stereoThis dataset contains 12,995 face images which are annotated with (1) five facial landmarks, (2) attributes of gender, smiling, wearing glasses, and head …
attribute, cnn, deep learning, detection, face, landmark detectionThe Pittsburgh Fast-food Image dataset (PFID) consists of 4545 still images, 606 stereo pairs, 3033600 videos for structure from motion, and 27 privacy-pr…
classification, food, laboratory, real, recognition, reconstruction, videoPlaces205 dataase contains 2.5 million images from 205 scene categories for the academic public. The image dataset contains 2,448,873 images from 205 sc…
feature, learning, place, recognition, scene, urbanThe CERTH image blur dataset consists of 2450 digital images, 1850 out of which are photographs captured by various camera models in different shooting co…
blur, defocus, detection, image, motion, qualityThe Zurich Summer v1.0 dataset is a collection of 20 chips (crops), taken from a QuickBird acquisition of the city of Zurich (Switzerland) in August 2002.…
aerial, annotation, city, gsd, nir, pan, rgb, satellite, segmentation, semantic, superpixel, urban, zurichThe Person Re-ID (PRID) 2011 dataset was created in co-operation with the Austrian Institute of Technology for the purpose of testing person re-identifica…
appearance, change, classification, graz, identification, illumination, multiview, pedestrian, trajectoryThis material is supplementary to Michael Stark, Bernt Schiele. How Good are Local Features for Classes of Geometric Objects. Eleventh IEEE Internatio…
binary, classification, object, shape, toolA New Color Image Database for Benchmarking of Face Detection Techniques and Human Skin Segmentation Techniques. A new color face image database for be…
benchmarking, detection, face, segmentation, skinThe Annotated Facial Landmarks in the Wild (AFLW) consists of a large-scale collection of annotated face images gathered from the web, exhibiting a large …
age, annotation, detection, face, landmark, poseThe Visual Attributes dataset contains visual attribute annotations for over 500 object classes (animate and inanimate) which are all represented in Image…
attribute, classification, imagenet, object, recognitionBackground Models Challenge (BMC) is a complete dataset and competition for the comparison of background subtraction algorithms. The main topics concern:…
background, change, detection, modeling, motion, segmentation, surveillance, videoThe Multi-FoV synthetic datasets are two synthetic scenes (vehicle moving in a city, and flying robot hovering in a confined room). For each scene, three …
blender, camera, fov, groundtruth, odometry, synthetic, visualThis data set comprises 144 images of an edge profile cutting head of a milling machine. The head tool contains a total of 30 cutting inserts. The cutting…
cutting, edge, head, inserts, localization, milling, monitoring, object, profile, tool, tools, wearOur repetitive pattern dataset with 106 images of app. 30 buildings from Pankrac, Prague and Marseille appearing in more than one image, number of appeara…
image classification, image retrieval, repetition, symmetry, urbanThe Dubrovnik6K and Rome16K datasets are image collections for SfM reconstruction, where the suffix refers to the number of images in the dataset. Dubro…
3d reconstruction, dubrovnik, landmark, rome, sfm, urbanMultispectral Imaging (MSI) datasets were acquired using IRIS II which is a lightweight portable system comprising of a high resolution camera, a novel fi…
alignment, groundtruth, illumination, matching, multi-spectral, registration, wavelengthThe CMP map2photo dataset consists of 6 pairs, where one image is satellite photo and second image is a map of the same area. The task is to match these …
baseline, description, detection, feature, map, matching, remote, sensing, wideThe PETS 2006 dataset contains 7 parts showing multi-sensor sequences containing left-luggage scenarios with increasing scene complexity at a train statio…
frontview, indoor, multitarget, object detection, object tracking, pedestrianThe GaTech VideoSeg dataset consists of two (waterski and yunakim?) video sequences for object segmentation. There exists no groundtruth segmentation an…
camera, model, motion, object, segmentation, video