Z
Zigang Geng
Researcher at University of Science and Technology of China
Publications - 6
Citations - 181
Zigang Geng is an academic researcher from University of Science and Technology of China. The author has contributed to research in topics: Computer science & Pixel. The author has an hindex of 3, co-authored 3 publications receiving 13 citations.
Papers
More filters
Proceedings ArticleDOI
Bottom-Up Human Pose Estimation Via Disentangled Keypoint Regression
TL;DR: In this paper, the authors proposed a disentangled keypoint regression (DEKR) method, which adopts adaptive convolutions through pixel-wise spatial transformer to activate the pixels in the keypoint regions and accordingly learn representations from them.
Journal ArticleDOI
Revealing the Dark Secrets of Masked Image Modeling
TL;DR: This paper compares MIM with the long-dominant supervised pre-trained models from two perspectives, the visualizations and the experiments, to uncover their key representational differences and finds that MIM models can perform significantly better on geometric and motion tasks with weak semantics or fine-grained classi-cation tasks, than their supervised counterparts.
Posted Content
Bottom-Up Human Pose Estimation by Ranking Heatmap-Guided Adaptive Keypoint Estimates.
TL;DR: A pixel-wise spatial transformer network is adopted to learn adaptive representations for handling the scale and orientation variance to further improve keypoint regression quality and a joint shape and heatvalue scoring scheme is presented to promote the estimated poses that are more likely to be true poses.
Journal ArticleDOI
All in Tokens: Unifying Output Space of Visual Tasks via Soft Token
TL;DR: AiT as discussed by the authors proposes a soft token to represent the task output, which can improve the accuracy of both the next token inference and decoding of the task outputs, and shows that a mask augmentation technique can greatly benefit these tasks.
Posted Content
Bottom-Up Human Pose Estimation Via Disentangled Keypoint Regression
TL;DR: In this article, the authors proposed a disentangled keypoint regression (DEKR) method, which adopts adaptive convolutions through pixel-wise spatial transformer to activate the pixels in the keypoint regions and accordingly learn representations from them.