Weakly Supervised Saliency Detection with A Category-Driven Map Generator
Kuang-Jui Hsu, Yen-Yu Lin and Yung-Yu Chuang
Abstract
Top-down saliency detection aims to highlight the regions of a specific object category, and typically relies on pixel-wise annotated training data. In this paper, we address
the high cost of collecting such training data by presenting a weakly supervised approach
to object saliency detection, where only image-level labels, indicating the presence or absence of a target object in an image, are available. The proposed framework is composed
of two deep modules, an image-level classifier and a pixel-level map generator. While
the former distinguishes images with objects of interest from the rest, the latter is learned
to generate saliency maps so that the training images masked by the maps can be better
predicted by the former. In addition to the top-down guidance from class labels, the map
generator is derived by also referring to other image information, including the background prior, area balance and spatial consensus. This information greatly regularizes
the training process and reduces the risk of overfitting, especially when learning deep
models with few training data.
Session
Posters
Files
Paper (PDF)
DOI
10.5244/C.31.67
https://dx.doi.org/10.5244/C.31.67
Citation
Kuang-Jui Hsu, Yen-Yu Lin and Yung-Yu Chuang. Weakly Supervised Saliency Detection with A Category-Driven Map Generator. In T.K. Kim, S. Zafeiriou, G. Brostow and K. Mikolajczyk, editors, Proceedings of the British Machine Vision Conference (BMVC), pages 67.1-67.13. BMVA Press, September 2017.
Bibtex
@inproceedings{BMVC2017_67,
title={Weakly Supervised Saliency Detection with A Category-Driven Map Generator},
author={Kuang-Jui Hsu, Yen-Yu Lin and Yung-Yu Chuang},
year={2017},
month={September},
pages={67.1-67.13},
articleno={67},
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.31.67},
isbn={1-901725-60-X},
url={https://dx.doi.org/10.5244/C.31.67}
}