Geppetto tracks 1035 robot training datasets including Open X-Embodiment, DROID, and AgiBot World, each with structured metadata on trajectory count, hours of demonstration, robot platforms, license, commercial use, access model, and file format. Every dataset connects to the robots it trains and the professions it affects.

DatasetYearCategoryTrajectoriesHoursPlatformsLicenseCommercialAccessFormat
DADA-20002019research-lab——dashcam accident videos with driver attentionnot statedNoopen—
DrivingStereo2019research-lab——car with stereo cameras and lidarnot statedNoopen—
EU Long-term Dataset2019research-lab——multi-sensor vehiclenot statedNoopen—
Fishyscapes2019research-lab——Cityscapes and LostAndFound images with anomaliesnot statedNoopen—
Honda H3D2019research-lab——HDD vehicle, 360 lidarresearch onlyNoopen—
Honda HEV-I2019research-lab——egocentric vehicle camera at intersectionsresearch onlyNoopen—
MulRan2019research-lab——car with radar and lidarregistration requiredNoopen—
nuScenes2019research-lab——car with 6 cameras, lidar, 5 radarsCC BY-NC-SA 4.0Nogated—
nuScenes, nuImages, nuPlan family2019research-lab——Motional vehiclesnot stated in fetched excerptNoopen—
SeeingThroughFog (DENSE)2019research-lab——car with gated camera, radar, lidar, FIRresearch and teachingNogated—
SemanticKITTI2019research-lab——KITTI odometry lidar sequencesCC BY-NC-SA 4.0Noopen—
Unsupervised LLAMAS2019research-lab——highway camera with lidar-map lane labelsnon-commercial researchNogated—
UrbanLoco2019research-lab——car with full sensor suite, San Francisco and Hong Kongnot statedNoopen—
Waymo Open Dataset2019research-lab——Waymo vehiclesWaymo terms (sign-in)Noopen—
Waymo Open Dataset Perception2019research-lab——Waymo vehicle, 5 lidars and 5 camerasWaymo non-commercial licenceNogated—
WoodScape2019research-lab——4 fisheye surround cameras, lidarnon-commercial researchNogated—
TEgO2019human-video——smartphone cameranot statedNoopen—
You2Me2019human-video——chest-mounted camera (per awesome list)not statedNoopen—
HO-3D2019human-video——one or several fixed RGB-D cameras (not egocentric)not statedNoopen—
ContactDB2019research-lab——human hand contact maps on objectsnot statedNoopen—
ClearGrasp2019research-lab——RGB-D of transparent objectsnot statedNoopen—
6-DOF GraspNet data2019research-lab——simulated parallel-jaw graspssee repo LICENSENoopen—
A Billion Ways to Grasp2019research-lab——simulated parallel-jaw grasps, dense SE(3) samplingnot statedNoopen—
RoboNet2019real-robot——fetch, franka_panda, kuka_iiwaCC-BY-4.0Yesopen—
ALFA2019research-lab—1.3fixed-wing UAVCC-BY-4.0Noopen—
Tsinghua Visual-Tactile Dataset (Multimodal Grasp Data Set)2019research-lab——not statednot statedNoopen—
AI4Animation human mocap sets (NSM 2019, Local Motion Phases 2020, Cranberry 2024)2019research-lab——human body mocapCC BY-NC 4.0Noopen—
NTU RGB+D 1202019research-lab——human body, Kinect RGB-D skeletonsnot statedNoopen—
PROX2019research-lab——human body in scanned 3D scenes, RGB-Dnot statedNoopen—
FreiHAND2019research-lab——RGB hand pose and shapenot statedNoopen—
ScanRefer2019research-lab——800 ScanNet scenesnot statedNoopen—
CVDN (Vision-and-Dialog Navigation)2019research-lab——Matterport3D simulatornot statedNoopen—
MP3D-EQA2019research-lab——Matterport3Dnot statedNoopen—
MT-EQA2019research-lab——simulated houses, multi-target questionsnot statedNoopen—
R4R (Room-for-Room)2019research-lab——Matterport3D simulatornot statedNoopen—
REVERIE2019research-lab——Matterport3D simulatornot statedNoopen—
Talk2Nav2019research-lab——Street Viewnot statedNoopen—
Talk2Car2019research-lab——nuScenes framesMITYesopen—
JRDB (JackRabbot Dataset and Benchmark)2019research-lab——JackRabbot social mobile manipulatorCC BY-NC-SA 3.0Nogated—
THOR2019research-lab——motion capture room with a mobile robot and 3D lidarnot statedNoopen—
ETH3D SLAM benchmark2019research-lab——not statednot statedNoopen—
Oxford Multimotion Dataset2019research-lab——static and dynamic sensorsnot statedNoopen—
OpenLORIS-Scene2019research-lab——service robotsnot statedNoopen—
UAVid2018research-lab——UAV 4K slanted-view videonot statedNoopen—
VisDrone2018research-lab——drone camerasnot statedNoopen—
Audi fleet CAN dataset (Drive2Vec)2018research-lab——Audi test vehicles with data loggersnot statedNoopen—
CULane2018research-lab——cameras on six vehicles, Beijingnon-commercialNoopen—
EuroCity Persons2018research-lab——car camera, 31 European citiesresearch useNoopen—
India Driving Dataset (IDD)2018research-lab——car front camera, Hyderabad and Bangalorenot statedNogated—
N-CARS2018research-lab——event camera car classificationnot statedNoopen—
PedX2018research-lab——stereo cameras and LiDAR at intersectionsnot statedNoopen—
EgoYouTubeHands / HandOverFace2018human-video——YouTube egocentric videosnot statedNoopen—
Mo2Cap22018human-video——cap-mounted fisheye cameranot statedNoopen—
Toybox2018human-video——Egocentric video of handheld object transformationsnot statedNoopen—
40K naturalistic 6-DoF grasp demonstrations2018research-lab——RightHand Robotics three-fingered ReFlex grippernot statedNoopen—
Robot Learning in Homes grasping dataset2018research-lab——low cost mobile robot arm in homesnot statedNoopen—
VMRD (Visual Manipulation Relationship Dataset)2018research-lab——RGB images of stacked objectsnot statedNoopen—
Yamaha-CMU Off-Road Dataset (YCOR)2018human-video———CC-BY-4.0Yesopen—
Jacquard2018research-lab——synthetic RGB-D with parallel-jaw grasp labelssigned EULANoopen—
More Than a Feeling2018research-lab——two-finger gripper with GelSight sensors on each fingernot statedNoopen—
Slip Detection with Combined Tactile and Visual Information2018research-lab——robot arm gripper with GelSight sensor and cameranot statedNoopen—
3DPW (3D Poses in the Wild)2018research-lab——human body, IMUs plus moving phone cameranot statedNoopen—
DeepMimic and AMP reference motion clips2018research-lab——simulated humanoid charactersnot statedNoopen—
LANI and CHAI2018research-lab——simulated 3D environmentsnot statedNoopen—
Talk the Walk2018research-lab——New York City 360 imagerynot statedNoopen—
Touchdown2018research-lab——Street View panoramasnot statedNoopen—
VirtualHome (ActivityPrograms)2018research-lab——Unity3D household simulatornot statedNoopen—
BDD-X2018research-lab——BDD dashcam videonot statedNoopen—
AI4Animation quadruped mocap (Mode-Adaptive Neural Networks)2018research-lab——dog mocapCC BY-NC 4.0Noopen—
MVSEC2018research-lab——stereo event cameras on car, motorbike, hexacopter, handheldnot statedNoopen—
ADVIO2018research-lab——iPhone, Tango, Pixelnot statedNoopen—
OxIOD2018research-lab——phone IMUsnot statedNoopen—
TUM VI Benchmark2018research-lab——handheld stereo camera and IMUnot statedNoopen—
AdobeIndoorNav2018research-lab——real indoor robot navigationnot statedNoopen—
Zurich Urban MAV Dataset2017research-lab——Fotokite MAV with GoPro, GPS, IMUno restrictionNoopen—
Mapillary Vistas2017research-lab——crowdsourced street-level camerasCC BY-NC-SA (per ad-datasets)Nogated—
CityPersons2017research-lab——Cityscapes pedestrian boxesnot statedNoopen—
Complex Urban (KAIST Urban)2017research-lab——car with lidars, stereo camera, GPS/IMU, altimeterCC BY-NC-SA 4.0Noopen—
DDD172017research-lab——DAVIS event camera on carnot statedNoopen—
NEXET2017research-lab——Nexar dashcam imagesnot statedNoopen—
Paris-Lille-3D2017research-lab——mobile LiDAR urban point cloudnot statedNoopen—
UDRIVE naturalistic driving database2017research-lab——instrumented cars, trucks and powered two-wheelersnot statedNoopen—
EgoDexter2017human-video——egocentric RGB-D sensornot statedNoopen—
Something-Something2017human-video——Crowd workers filming their own handsnot statedNoopen—
THU-READ2017human-video——RGB-D sensor mounted on a helmetnot statedNoopen—
Dex-Net 2.02017research-lab——synthetic point clouds, parallel-jaw graspsnot statedNoopen—
Wire Detection Dataset2017human-video——droneCC-BY-4.0Yesopen—
Something-Something V22017research-lab——human hands with everyday objectsotherNoopen—
EgoGesture2017human-video——head-mounted RGB-D cameranot statedNoopen—
Integrated grasping and vision simulator dataset (Veres)2017research-lab——simulated multi-fingered dexterous handnot statedNoopen—
MIT-Princeton ARC 2017 grasping and suction data2017research-lab——robot picking system, RGB-Dnot statedNoopen—
The Feeling of Success2017research-lab——two-finger gripper with GelSight sensors on each fingernot statedNoopen—
TotalCapture2017research-lab——human body, multi-view video plus IMU plus Viconnot statedNoopen—
EQA (Embodied Question Answering)2017research-lab——simulated housesnot statedNoopen—
IQUAD V12017research-lab——AI2-THORnot statedNoopen—
R2R (Room-to-Room)2017research-lab——Matterport3D simulatornot statedNoopen—
Aachen Day-Night, RobotCar Seasons, CMU Seasons2017research-lab——outdoor visual localizationnot statedNoopen—
PennCOSYVIO2017research-lab——handheld visual-inertial rignot statedNoopen—
Active Vision Dataset2017research-lab——robot-captured dense RGB-Dnot statedNoopen—
L-CAS 3D Point Cloud People Dataset2017research-lab——3D LiDAR on mobile robotnot statedNoopen—