AMR decouples decoder and uses adaptive guidance on high-motion regions to accelerate masked skeleton reconstruction pre-training and raise action recognition accuracy on NTU RGB+D and PKU-MMD datasets.
Hyperbolic self-paced learning for self-supervised skeleton-based action representations
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CV 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
A Lorentz-model hyperbolic framework for semantic segmentation that integrates with Euclidean networks, provides free uncertainty maps, and is validated on ADE20K, COCO-Stuff, Pascal-VOC and Cityscapes using DeepLabV3, SegFormer, Mask2Former and MaskFormer.
citing papers explorer
-
Exploring Adaptive Masked Reconstruction for Self-Supervised Skeleton-Based Action Recognition
AMR decouples decoder and uses adaptive guidance on high-motion regions to accelerate masked skeleton reconstruction pre-training and raise action recognition accuracy on NTU RGB+D and PKU-MMD datasets.
-
Lorentz Framework for Semantic Segmentation
A Lorentz-model hyperbolic framework for semantic segmentation that integrates with Euclidean networks, provides free uncertainty maps, and is validated on ADE20K, COCO-Stuff, Pascal-VOC and Cityscapes using DeepLabV3, SegFormer, Mask2Former and MaskFormer.