| Name/Description | size | images/videos | train | val | test | Download |
|---|---|---|---|---|---|---|
| Human Parsing detailed pixel-wise annotations for fashion images |
BaiduDisk(kjgk) | |||||
| [LIP](https://vuhcs.github.io |
10K | |||||
Chictopia Plus A Generative Model of People in Clothing |
2.99G | 14400 | ||||
| Look into Person(LIP) elaborated pixel-wise annotations with 19 semantic human part labels and 2D human poses with 16 key points, cropped person instances from COCO dataset with size larger than 50 * 50 |
50462 | 30462 | 10000 | 10000 | GoogleDrive BaiduDisk | |
| LIP-4-1:CIHP Crowd Instance-level Human Parsing Dataset |
38280 | 28280 | 5000 | 5000 | GoogleDrive BaiduDisk | |
| LIP-4-2:VIP Video instance-level Parsing For every 25 consecutive frames in each video, one frame is annotated densely with pixel-wise semantic part categories and instance-level identification. There are 21247 densely annotated images in total. |
404 | 304 | 50 | 50 | OneDrive BaiduDisk | |
| LIP-4-5:DCPS Dark Complexion Portrait Segmentation Dataset |
14735 | 12165 | 2570 | GoogleDrive BaiduDisk(fh52) | ||
Clothing Co-Parsing |
3K+ | |||||
| http://robotics.dei.unipd.it/reid/index.php |
||||||
BackgroundMattingV2:VideoMatte240K contains 484 pairs of high-resolution alpha matte and foreground video clips, constituting 240,709 unique frames. The alpha matte and foreground data were extracted from green screen stock footage purchased by the University of Washington.384 4K+100 HD |
4.7G | 484 | GoogleDrive BaiduDisk | |||
BackgroundMattingV2:PhotoMatte85 85 protrait images |
395M | 85 | GoogleDrive BaiduDisk | |||
| YouTubeVOS-2019 720p, annotated: 6fps, raw images: 30fps, 94 object categories |
4000+ | GoogleDrive BaiduDisk(uu4q) | ||||
| YouTubeVIS-2021 |
12.3G | 3859 | 2985 | 421 | 453 | GoogleDrive BaiduDisk(fwc5) |
3,859 high-resolution YouTube videos, 2,985 training videos, 421 validation videos and 453 test videos. An improved 40-category label set by merging eagle and owl into bird, ape into monkey, deleting hands, and adding flying disc, squirrel and whale 8,171 unique video instances 232k high-quality manual annotations
CV Datasets on the web
awesome-segmentation-saliency-dataset
pixabay.com - search transparent images