AutoScaler: Scale-Attention Networks for Visual Correspondence
Shenlong Wang, Linjie Luo, Ning Zhang and Jia Li
Abstract
Finding visual correspondence between local features is key to many computer vision
problems. While defining features with larger contextual scales usually implies greater
discriminativeness, it could also lead to less spatial accuracy of the features. We propose AutoScaler, a scale-attention network to explicitly optimize this trade-off in visual
correspondence tasks. Our architecture consists of a weight-sharing feature network to
compute multi-scale feature maps and an attention network to combine them optimally
in the scale space. This allows our network to have adaptive sizes of equivalent receptive field over different scales of the input. The entire network can be trained end-to-end
in a Siamese framework for visual correspondence tasks. Using the latest off-the-shelf
architecture for the feature network, our method achieves competitive results compared
to state-of-the-art methods on challenging optical flow and semantic matching benchmarks, including Sintel, KITTI and CUB-2011. We also show that our attention network
alone can be applied to existing hand-crafted feature descriptors (e.g Daisy) and improve
their performance on visual correspondence tasks.
Session
Orals - Matching
Files
Paper (PDF)
DOI
10.5244/C.31.185
https://dx.doi.org/10.5244/C.31.185
Citation
Shenlong Wang, Linjie Luo, Ning Zhang and Jia Li. AutoScaler: Scale-Attention Networks for Visual Correspondence. In T.K. Kim, S. Zafeiriou, G. Brostow and K. Mikolajczyk, editors, Proceedings of the British Machine Vision Conference (BMVC), pages 185.1-185.13. BMVA Press, September 2017.
Bibtex
@inproceedings{BMVC2017_185,
title={AutoScaler: Scale-Attention Networks for Visual Correspondence},
author={Shenlong Wang, Linjie Luo, Ning Zhang and Jia Li},
year={2017},
month={September},
pages={185.1-185.13},
articleno={185},
numpages={13},
booktitle={Proceedings of the British Machine Vision Conference (BMVC)},
publisher={BMVA Press},
editor={Tae-Kyun Kim, Stefanos Zafeiriou, Gabriel Brostow and Krystian Mikolajczyk},
doi={10.5244/C.31.185},
isbn={1-901725-60-X},
url={https://dx.doi.org/10.5244/C.31.185}
}