TY - GEN
T1 - Contour knowledge transfer for salient object detection
AU - Li, Xin
AU - Yang, Fan
AU - Cheng, Hong
AU - Liu, Wei
AU - Shen, Dinggang
N1 - Funding Information:
Acknowledgments. This research was funded in part by the National Key R&D Program of China (2017YFB1302300), the National Nature Science Foundation of China (U1613223), and the Open Research Subject of Comprehensive Health Management Center of Xihua University (JKGL2018-029).
Funding Information:
This research was funded in part by the National Key R&D Program of China (2017YFB1302300), the National Nature Science Foundation of China (U1613223), and the Open Research Subject of Comprehensive Health Management Center of Xihua University (JKGL2018-029).
Publisher Copyright:
© Springer Nature Switzerland AG 2018.
PY - 2018
Y1 - 2018
N2 - In recent years, deep Convolutional Neural Networks (CNNs) have broken all records in salient object detection. However, training such a deep model requires a large amount of manual annotations. Our goal is to overcome this limitation by automatically converting an existing deep contour detection model into a salient object detection model without using any manual salient object masks. For this purpose, we have created a deep network architecture, namely Contour-to-Saliency Network (C2S-Net), by grafting a new branch onto a well-trained contour detection network. Therefore, our C2S-Net has two branches for performing two different tasks: (1) predicting contours with the original contour branch, and (2) estimating per-pixel saliency score of each image with the newly-added saliency branch. To bridge the gap between these two tasks, we further propose a contour-to-saliency transferring method to automatically generate salient object masks which can be used to train the saliency branch from outputs of the contour branch. Finally, we introduce a novel alternating training pipeline to gradually update the network parameters. In this scheme, the contour branch generates saliency masks for training the saliency branch, while the saliency branch, in turn, feeds back saliency knowledge in the form of saliency-aware contour labels, for fine-tuning the contour branch. The proposed method achieves state-of-the-art performance on five well-known benchmarks, outperforming existing fully supervised methods while also maintaining high efficiency.
AB - In recent years, deep Convolutional Neural Networks (CNNs) have broken all records in salient object detection. However, training such a deep model requires a large amount of manual annotations. Our goal is to overcome this limitation by automatically converting an existing deep contour detection model into a salient object detection model without using any manual salient object masks. For this purpose, we have created a deep network architecture, namely Contour-to-Saliency Network (C2S-Net), by grafting a new branch onto a well-trained contour detection network. Therefore, our C2S-Net has two branches for performing two different tasks: (1) predicting contours with the original contour branch, and (2) estimating per-pixel saliency score of each image with the newly-added saliency branch. To bridge the gap between these two tasks, we further propose a contour-to-saliency transferring method to automatically generate salient object masks which can be used to train the saliency branch from outputs of the contour branch. Finally, we introduce a novel alternating training pipeline to gradually update the network parameters. In this scheme, the contour branch generates saliency masks for training the saliency branch, while the saliency branch, in turn, feeds back saliency knowledge in the form of saliency-aware contour labels, for fine-tuning the contour branch. The proposed method achieves state-of-the-art performance on five well-known benchmarks, outperforming existing fully supervised methods while also maintaining high efficiency.
KW - Deep learning
KW - Saliency detection
KW - Transfer learning
UR - http://www.scopus.com/inward/record.url?scp=85055424049&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-01267-0_22
DO - 10.1007/978-3-030-01267-0_22
M3 - Conference contribution
AN - SCOPUS:85055424049
SN - 9783030012663
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 370
EP - 385
BT - Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings
A2 - Weiss, Yair
A2 - Ferrari, Vittorio
A2 - Sminchisescu, Cristian
A2 - Hebert, Martial
PB - Springer Verlag
T2 - 15th European Conference on Computer Vision, ECCV 2018
Y2 - 8 September 2018 through 14 September 2018
ER -