JHMDB

Joint-annotated Human Motion Data Base

JHMDB is an action recognition dataset that consists of 960 video sequences belonging to 21 actions. It is a subset of the larger HMDB51 dataset collected from digitized movies and YouTube videos. The dataset contains video and annotation for puppet flow per frame (approximated optimal flow on the person), puppet mask per frame, joint positions per frame, action label per clip and meta label per clip (camera motion, visible body parts, camera viewpoint, number of people, video quality). Source: Unsupervised Deep Metric Learning via Orthogonality based Probabilistic Loss Image Source: https://arxiv.org/pdf/1712.06316.pdf

Videos

벤치마크

Skeleton Based Action Recognition on J-HMDB 결과 104개

Skeleton Based Action Recognition on JHMDB (2D poses only) 결과 48개

Referring Expression Segmentation on J-HMDB 결과 42개

Skeleton Based Action Recognition on JHMDB Pose Tracking 결과 24개

Action Detection on J-HMDB 결과 18개

Skeleton Based Action Recognition on J-HMBD Early Action 결과 16개

Pose Estimation on J-HMDB 결과 15개

2D Human Pose Estimation on JHMDB (2D poses only) 결과 5개

Open Vocabulary Action Detection on JHMDB 결과 1개