Event-Datasets.md

Dataset	Description	Tasks	Cameras	Metric
HQF	Events and ground truth frames	Image Reconstruction	DAVIS240C	-
DR	Indoor and outdoor environment	Image Reconstruction	miniDAVIS346	-
DAVIS240C	Events and ground-truth, intensity frames and IMU	Image Reconstruction and SLAM	DAVIS+Simulator	SSIM
DHP19	33 movements recorded from 17 subjects	3D Human Pose Estimation	DVS cameras	MPJPE
EventCap	12 sequences of 6 actors performing different activities	3D Human Pose Estimation	DAVIS240C	MPJPE
DDD17	12 hours of driving data	Semantic Segmentation	DAVIS	mIOU
DDD20	51h of DAVIS event+frame camera and vehicle human control data, collected from 4000 km of highway and urban driving	Semantic Segmentation	DAVIS	mIOU
DSEC-Semantic	53 driving sequences collected in a variety of urban and rural environments in Switzerland	Semantic Segmentation	DAVIS	mIOU
EVIMO2	3.75 minutes of independently moving household objects, 22.55 minutes of static scenes, and 14.85 minutes of basic motions in shallow scenes	Semantic Segmentation	Prophesee Gen3 VGA	mIOU
KITTI Simulated	20 videos for object tracking, converted from videos to events using the V2E simulator	Object Detection	Simulator	mAP
Gen1 Automative	226719 cars and 27658 pedestrians with manual bounding box annotations	Object Detection	ATIS	mAP
1Mpx Detection	event streams and more than 25 million bounding boxes of cars, pedestrians, and two-wheelers	Object Detection	1 megapixel event camera+GoPro Hero6	mAP
PKU-DDD17-CAR	discontinuous images and event streams of driving scenes	Object Detection	DAVIS346	mAP
N-Cars	4,029 event samples for the binary task of car recognition	Object Recognition	ATIS	Accuracy
N-Caltech101	8,246 samples and 100 classes, based on Caltech101 dataset	Object Recognition	ATIS	Accuracy
ENFS-real	First real-world dataset involving multi-scale LR-HR pairs for event SR	Super-resolution	AUO80e+DAVIS34	RMSE
ENFS-syn	Involves 2(4, 8, 16)× LR-HR pairs for 65 scenes	Super-resolution	Event Simulator	RMSE
RGB-DAVIS-syn	Similar to ENFS-syn, utilizing the HR frames from RGB-DAVIS	Super-resolution	Event Simulator	RMSE
DVSNOISE20	contains 16 different scenes mostly under static conditions	Denoising & Super-resolution	DAVIS346	IoU
DVSCLEAN	First real dataset shares the same recording method with simulated dataset	Denoising	CeleX-V	SNR
DND21	The only one based on clean DVS plus realistic DVS leak and shot noise	Denoising	DAVIS346+Event Simulator	AUC
IEBCS	An extended DVS pixel simulator simplifying the latency and the noise models	Denoising	Event Simulator	Accuracy
v2e	Including pixel level Gaussian event threshold mismatch, finite intensity dependent bandwidth, and intensity-dependent noise	Denoising	Event Simulator	Accuracy
ESIM	Multiple synthetic events and APS images	Multiple synthetic events and APS images	Image Super-resolution	Event Simulator
MVSEC	consists of day/night-time outdoor driving sequences and flying indoor sequences, with depth GT	Depth Estimation	DAVIS	AAE / MAE
DESC	contains 53 driving scenarios taken in various lighting conditions	Depth Estimation	Stereo RGB and event camera	MAE / RMSE
VIVID	contains sequences for visual navigation in poor illumination condition	Depth Estimation	DAVIS 240 C	MAE / RMSE
Vector	benchmark datasets for research on multi-sensor SLAM including data from an event stereo camera, a stereo camera, an RGB-D sensor, a LiDAR, and an IMU	Visual Odometry and SLAM	Prophesee Gen3 VGA	-
TMU-VIE	benchmark datasets for developing 3D perception and navigation algorithms, including stereo events and grayscale frames, IMU and 6dof motion data	Visual Odometry and SLAM	Prophesee Gen4 HD	-
EDS	A dataset released for direct monocular visual odometry, including data from an event camera and a RGB camera	Visual Odometry and SLAM	Prophesee EVALUATION
Stereo DAVIS	Stereo events data and their corresponding grayscale frames, an IMU data	Visual Odometry and SLAM	DAVIS+Simulator	-