Robotics and Computer Vision Lab

Publications

조회 수 1540 댓글 0
Extra Form
저 자 Tae-Hyun Oh, Yasuyuki Matsushita, Yu-Wing Tai, In So Kweon
학 회 IEEE International Conference on Computer Vision and Pattern Recognition (CVPR)
Notes <iframe width="560" height="315" src="https://www.youtube.com/embed/PzE31ZpG4_I" frameborder="0" allowfullscreen></iframe>
논문일시(Year) 2015
논문일시(Month) 06
http://thohkaistackr.wix.com/page#!projecttpami15/ca3vRank minimization problem can be boiled down to either Nuclear Norm Minimization (NNM) or Weighted NNM (WNNM) problem. The problems related to NNM (or WNNM) can be solved iteratively by applying a closedform proximal operator, called Singular Value Thresholding (SVT) (or Weighted SVT), but they suffer from high computational cost to compute a Singular Value Decomposition (SVD) at each iteration. In this paper, we propose an accurate and fast approximation method for SVT, called fast randomized SVT (FRSVT), where we avoid direct computation of SVD. The key idea is to extract an approximate basis for the range of a matrix from its compressed matrix.
Given the basis, we compute the partial singular values of the original matrix from a small factored matrix. While the basis approximation is the bottleneck, our method is already severalfold faster than thin SVD. By adopting a range propagation technique, we can further avoid one of the bottleneck at each iteration. Our theoretical analysis provides a stepping stone between the approximation bound of SVD
and its effect to NNM via SVT. Along with the analysis, our empirical results on both quantitative and qualitative studies show our approximation rarely harms the convergence behavior of the host algorithms. We apply it and validate the efficiency of our method on various vision problems, e.g. subspace clustering, weather artifact removal, simultaneous multi-image alignment and rectification.


[Project page]
Matlab souce code is available.

[Acknowledgment (NCRC)];
The first author was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (No. 2010-0028680).

[Award]
* This work was awarded by Samsung HumanTech Paper competition as Gold Prize (1st place).

List of Articles
432. Multi-scale Pyramid Pooling for Deep Convolutional Representation
Donggeun Yoo, Sunggyun Park, Joon-Young Lee, In So Kweon
IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPR Workshop - DeepVision) 2015 / 06
431. One-day outdoor photometric stereo via skylight estimation
Jiyoung Jung, Joon-Young Lee, In So Kweon
IEEE Conference on Computer Vision and Pattern Recognition (CVPR) [oral] 2015 / 06
430. 하늘빛 분포 추정을 이용한 야외 광도 스테레오
Jiyoung Jung, Joon-Young Lee, In So Kweon
영상처리 및 이해에 관한 워크샵 (IPIU) 2015 / 02
429. Accurate Depth Map Estimation from a Lenslet Light Field Camera
Hae-Gon Jeon, Jaesik Park, Gyeongmin Choe, Jinsun Park, Yunsu Bok, Yu-Wing Tai, In So Kweon
IEEE International Conference on Computer Vision and Pattern Recognition (CVPR) 2015 / 06
» Fast Randomized Singular Value Thresholding for Nuclear Norm Minimization
Tae-Hyun Oh, Yasuyuki Matsushita, Yu-Wing Tai, In So Kweon
IEEE International Conference on Computer Vision and Pattern Recognition (CVPR) 2015 / 06
427. Multispectral Pedestrian Detection: Benchmark Dataset and Baselines
Soonmin Hwang, Jaesik Park, Namil Kim, Yukyung Choi, In So Kweon
IEEE International Conference on Computer Vision and Pattern Recognition (CVPR) 2015 / 06
426. 깊이 영상 처리를 위한 학습기반 신뢰도 추정 및 재질 분류
Min-Hyun Kim, Jaesik Park, In So Kweon
제 27회 영상처리 및 이해에 관한 워크샵 (IPIU) 2015 / 02
425. 잔디 환경에서 단일 카메라와 보조 감시카메라를 이용한 로봇 자세 추정
Kibaek Park, Seunghak Shin, In So Kweon
제 27회 영상처리 및 이해에 관한 워크샵 (IPIU) 2015 / 02
424. 음악 인터페이스를 위한 깊이 영상 기반의 속도 감응형 터치 센서
Hyungil Kim, Woontack Woo, Gyeongmin Choe, In So Kweon
제 27회 영상처리 및 이해에 관한 워크샵 (IPIU) 2015 / 02
423. 컬러-열영상 퓨전을 통한 강인한 보행자 검출 기법
Soonmin Hwang, Jaesik Park, Namil Kim, Yukyung Choi, In So Kweon
제 27회 영상처리 및 이해에 관한 워크샵 2015 / 02
422. 강인한 보행자 검출을 위한 Plug-in 모델
Soonmin Hwang, Tae-Hyun Oh, In So Kweon
The 9th Korea Robotics Society Annual Conference (KRoC 2014) 2014 / 06
421. 시야 공유가 없는 다중 카메라를 이용한 사람 추적
Junsik Kim, Kyungdon Joo, Tae-Hyun Oh, Jaesik Park, In So Kweon
제 27회 영상처리 및 이해에 관한 워크샵 (IPIU) 2015 / 02
420. An Autonomous Driving System for Unknown Environments using a Unified Map
Inwook Shim, Jongwon Choi, Seunghak Shin, Tae-Hyun Oh, Unghui Lee, Byungtae Ahn, Dong-Geol Choi, David Hyunchul Shim, In So Kweon
IEEE Transactions on Intelligent Transportation Systems (T-ITS) 2015 / 08
419. Specular Reflection Separation Using Dark Channel Prior
Hyeongwoo Kim, Hailin Jin, Sunil Hadap, In So Kweon
IEEE International Conference on Computer Vision and Pattern Recognition (CVPR) 2013 / 06
418. Hybrid vision-based SLAM coupled with moving object tracking
Jihong Min, Jungho Kim, Hyeongwoo Kim, Kiho Kwak, In So Kweon
International Conference on Robotics & Automation (ICRA) 2014 / 05
417. Sensor Fusion of Cameras and a Laser for City-Scale 3D Reconstruction
Yunsu Bok, Dong-Geol Choi, In So Kweon
Sensors 2014 / 11
416. A Two Phase Approach for Pedestrian Detection
Soonmin Hwang, Tae-Hyun Oh, In So Kweon
The 12th Asian Conference on Computer Vision Workshops (ACCV Workshops) 2014 / 11
415. Time-of-flight sensor calibration for a color and depth camera pair
Jiyoung Jung, Joon-Young Lee, Yekeun Jeong, In So Kweon
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 2015 / 07
414. High Quality Depth Map Upsampling and Completion for RGB-D cameras
Jaesik Park, Hyeongwoo Kim, Yu-Wing Tai, Michael S. Brown, In So Kweon
IEEE Transactions on Image Processing (TIP) 2014 / 09
413. Robust High Dynamic Range Imaging by Rank Minimization
Tae-Hyun Oh, Joon-Young Lee, Yu-Wing Tai, In So Kweon
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 2015 / 06
Board Pagination Prev 1 ... 5 6 7 8 9 10 11 12 13 14 ... 31 Next
/ 31