EgoPressDiff is a multimodal video diffusion framework that generates UV-pressure maps from egocentric visual inputs using PoseNet, Vertex Encoder, and a Distribution-Calibrated Spatial Layer, reporting over 34% relative Volumetric IoU improvement on the EgoPressure dataset.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
EgoPressDiff: Multimodal Video Diffusion for Egocentric UV-Domain Hand-Pressure Estimation
EgoPressDiff is a multimodal video diffusion framework that generates UV-pressure maps from egocentric visual inputs using PoseNet, Vertex Encoder, and a Distribution-Calibrated Spatial Layer, reporting over 34% relative Volumetric IoU improvement on the EgoPressure dataset.