[2003.12230] Learning to Optimize Non-Rigid Tracking

One of the widespread solutions for non-rigid tracking has a nested-loop structure: with Gauss-Newton to minimize a tracking objective in the outer loop, and Preconditioned Conjugate Gradient (PCG) to solve a sparse linear system in the inner loop. In this paper, we employ learnable optimizations to improve tracking robustness and speed up solver convergence. First, we upgrade the tracking objective by integrating an alignment data term on deep features which are learned end-to-end through CNN. The new tracking objective can capture the global deformation which helps Gauss-Newton to jump over local minimum, leading to robust tracking on large non-rigid motions. Second, we bridge the gap between the preconditioning technique and learning method by introducing a ConditionNet which is trained to generate a preconditioner such that PCG can converge within a small number of steps. Experimental results indicate that the proposed learning method converges faster than the original PCG by a lar

1 mentions: @sam_murayama
Date: 2021/02/20 14:21

Referring Tweets

@sam_murayama Non-Rigid Tracking Simultaneous Tracking on Mapping 非剛体のトラッキングが可能になれば、非剛体を含む物体のone shotだけでなく、それをトラッキングすることにより、大量の教師データを得ることができる。 t.co/EMPXD44YT0

Related Entries

Read more 8-8 分析の評価: 標準誤差 (2): Introduction to statistical data analysis Ⅱ - 8 (Yuta Koike, UTokyo OE)
0 users, 2 mentions 2020/12/27 12:52
Read more 8-10 分析の評価: t値とp値 (1): Introduction to statistical data analysis Ⅱ - 8 (Yuta Koike, UTokyo OE)
0 users, 1 mentions 2020/12/27 14:21
Read more 9-1 重回帰分析: 復習: Introduction to statistical data analysis Ⅱ - 9 (Yuta Koike, UTokyo OE)
0 users, 1 mentions 2020/12/31 02:21
Read more セマンティックセグメンテーション [初期の手法(FCN/SegNet/U-Net)から,PSPNet, DeepLab v3 まで] | CVMLエキスパートガイド
0 users, 4 mentions 2021/01/04 02:21
Read more Zero-shot learningの紹介:見たことがない画像やニュースを予測してみました | GMOインターネット 次世代システム研究室
0 users, 1 mentions 2021/03/01 06:54