UVO is a new benchmark for open-world class-agnostic object segmentation in videos.
23 PAPERS • 3 BENCHMARKS
…We benchmark four foundational video understanding tasks: action recognition, action segmentation, object detection and multi-object tracking.
1 PAPER • NO BENCHMARKS YET
…Segmentation masks Bounding boxes For the full description of labels and metadata, check out the README.
0 PAPER • NO BENCHMARKS YET