FMI Data Index

The FMI Data Index is an open repository for machine learning datasets created by volunteer students from the Faculty of Mathematics and Computer Science, University of Bucharest. The repository is indended to be used for both research and teaching purposes.

All information compiled in the webpages available in the table below, including images and annotations, comes from the official pages of the datasets.

Dataset Name Dataset Type
Agriculture-Vision CV
AU-AIR CV
COCO: Common Objects in Context CV
DensePose: Dense Human Pose Estimation In The Wild CV
FaceForensics++: Learning to Detect Manipulated Facial Images CV
The German Traffic Sign Benchmark CV
Google Landmarks Dataset V2 CV
GOT-10k: Generic Object Tracking Benchmark CV
The Intersection Drone Dataset CV
The Unsupervised Labeled Lane Markers Dataset CV
Logo-2K+:A Large-Scale Logo Dataset for Scalable Logo Classification CV
LVIS: A Dataset For Large Vocabulary Instance Segmentation CV
Open Images Dataset V6 CV
PANDA: A Gigapixel-level Human-centric Video Dataset CV
PDESTRE: Fully Annotated Datasets for Pedestrian Detection,
Tracking, Re-Identification and Search from Aerial Devices		 CV
PedX CV
QMUL-OpenLogo: Open Logo Detection Challenge Dataset CV
RoadText-1K: Text Detection & Recognition Dataset for Driving Videos CV
The Semantic Drone Dataset CV
TACO: Trash Annotations in Context CV
Vehicle Make and Model Recognition Dataset CV
Vision Meets Drone Challenge Dataset CV
WiderPerson: A Diverse Dataset for Dense Pedestrian Detection in the Wild CV
Computer Vision for Building Damage Assessment Dataset CV