A wide range of approaches have been developed for camera pose estimation, including correspondence-based methods, end-to-end pose regression, and recent 3D geometric foundation models. Our key observation is that no single estimator is optimal for diverse challenges, such as wide baselines, lack of texture, appearance...
Zhi-Ning Gu, Shang-Jie Du, Wei-Min Qiu et al.· 0 citations
Video generation models have demonstrated emerging zero-shot capabilities for visual reasoning, perception, and other vision tasks. However, diffusion-based video generation is inherently stochastic, while many downstream vision tasks are deterministic. Motivated by the effectiveness of self-consistency in chain-of-tho...
GramLoop is introduced, a training-free framework that replays a short transformer window and controls each replay through final-layer cosine-Gram consistency, which improves object detection and semantic segmentation under corruptions, perturbations, and natural shifts.
Yang Chen, Can-Yu Shen, Xin-Zhe Rao et al.· 0 citations
A training-free looping framework that repeatedly applies selected transformer layers inside each denoising call is introduced, which improves primary and auxiliary quality metrics with competitive quality--efficiency trade-offs across two Scale-RAE model scales.
Yuan-Yi Yan, Xin-Zhe Rao, Can-Yu Shen et al.· 0 citations
OODA-Tool, a typed closed-loop policy designed to mitigate state preservation from action realization, consistently improves task success across model sizes, with larger gains on smaller models and on tasks whose actions depend strongly on information accumulated across turns and prior tool results.
Rong-Feng Guo, Yin-Xuan Huang, Yusen Wu et al.· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
