WTPose combines a Swin transformer backbone with a multi-scale waterfall module using dilated neighborhood attention and reports +1.2 AP over Swin-B on COCO val.
Realtime multi-person 2d pose estimation us- ing part affinity fields
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Waterfall Transformer for Multi-person Pose Estimation
WTPose combines a Swin transformer backbone with a multi-scale waterfall module using dilated neighborhood attention and reports +1.2 AP over Swin-B on COCO val.