Zobrazeno 1 - 10
of 111
pro vyhledávání: '"Chen, Liangjian"'
We present Position Interpolation (PI) that extends the context window sizes of RoPE-based pretrained LLMs such as LLaMA models to up to 32768 with minimal fine-tuning (within 1000 steps), while demonstrating strong empirical results on various tasks
Externí odkaz:
http://arxiv.org/abs/2306.15595
Autor:
Kong, Deying, Zhang, Linguang, Chen, Liangjian, Ma, Haoyu, Yan, Xiangyi, Sun, Shanlin, Liu, Xingwei, Han, Kun, Xie, Xiaohui
Reconstructing 3D hand meshes from monocular RGB images has attracted increasing amount of attention due to its enormous potential applications in the field of AR/VR. Most state-of-the-art methods attempt to tackle this task in an anonymous manner. S
Externí odkaz:
http://arxiv.org/abs/2209.10840
Autor:
Ma, Haoyu, Wang, Zhe, Chen, Yifei, Kong, Deying, Chen, Liangjian, Liu, Xingwei, Yan, Xiangyi, Tang, Hao, Xie, Xiaohui
Recently, the vision transformer and its variants have played an increasingly important role in both monocular and multi-view human pose estimation. Considering image patches as tokens, transformers can model the global dependencies within the entire
Externí odkaz:
http://arxiv.org/abs/2209.08194
Publikováno v:
In International Immunopharmacology 5 December 2024 142 Part B
Autor:
Ma, Haoyu, Chen, Liangjian, Kong, Deying, Wang, Zhe, Liu, Xingwei, Tang, Hao, Yan, Xiangyi, Xie, Yusheng, Lin, Shih-Yao, Xie, Xiaohui
Estimating the 2D human poses in each view is typically the first step in calibrated multi-view 3D pose estimation. But the performance of 2D pose detectors suffers from challenging situations such as occlusions and oblique viewing angles. To address
Externí odkaz:
http://arxiv.org/abs/2110.09554
Publikováno v:
2021 IEEE Winter Conference on Applications of Computer Vision (WACV)
Estimating 3D hand pose directly from RGB imagesis challenging but has gained steady progress recently bytraining deep models with annotated 3D poses. Howeverannotating 3D poses is difficult and as such only a few 3Dhand pose datasets are available,
Externí odkaz:
http://arxiv.org/abs/2012.03205
Publikováno v:
2021 IEEE Winter Conference on Applications of Computer Vision (WACV)
Estimating 3D hand poses from a single RGB image is challenging because depth ambiguity leads the problem ill-posed. Training hand pose estimators with 3D hand mesh annotations and multi-view images often results in significant performance gains. How
Externí odkaz:
http://arxiv.org/abs/2012.03206
Publikováno v:
2020 IEEE Winter Conference on Applications of Computer Vision (WACV)
Estimating3D hand poses from RGB images is essentialto a wide range of potential applications, but is challengingowing to substantial ambiguity in the inference of depth in-formation from RGB images. State-of-the-art estimators ad-dress this problem
Externí odkaz:
http://arxiv.org/abs/2012.03197
Autor:
Wu, Zhenyu, Hoang, Duc, Lin, Shih-Yao, Xie, Yusheng, Chen, Liangjian, Lin, Yen-Yu, Wang, Zhangyang, Fan, Wei
Estimating the 3D hand pose from a monocular RGB image is important but challenging. A solution is training on large-scale RGB hand images with accurate 3D hand keypoint annotations. However, it is too expensive in practice. Instead, we have develope
Externí odkaz:
http://arxiv.org/abs/2010.01158
Akademický článek
Tento výsledek nelze pro nepřihlášené uživatele zobrazit.
K zobrazení výsledku je třeba se přihlásit.
K zobrazení výsledku je třeba se přihlásit.