Learning Deep Structured Multi-Scale Features using Attention-Gated CRFs for Contour Prediction

Dan Xu (1) , Wanli Ouyang (2) , Xavier Alameda-Pineda (3) , Elisa Ricci (4) , Xiaogang Wang (5) , Nicu Sebe (1)

1 UNITN - Università degli Studi di Trento = University of Trento
2 The University of Sydney
3 PERCEPTION - Interpretation and Modelling of Images and Videos
4 FBK - Fondazione Bruno Kessler [Trento, Italy]
5 CUHK - The Chinese University of Hong Kong [Hong Kong]

Recent works have shown that exploiting multi-scale representations deeply learned via convolutional neural networks (CNN) is of tremendous importance for accurate contour detection. This paper presents a novel approach for predicting contours which advances the state of the art in two fundamental aspects, i.e. multi-scale feature generation and fusion. Different from previous works directly considering multi-scale feature maps obtained from the inner layers of a primary CNN architecture, we introduce a hierarchical deep model which produces more rich and complementary representations. Furthermore, to refine and robustly fuse the representations learned at different scales, the novel Attention-Gated Conditional Random Fields (AG-CRFs) are proposed. The experiments ran on two publicly available datasets (BSDS500 and NYUDv2) demonstrate the effectiveness of the latent AG-CRF model and of the overall hierarchical framework.

Domaines: Informatique [cs]

Dan Xu, Wanli Ouyang, Xavier Alameda-Pineda, Elisa Ricci, Xiaogang Wang, et al.. Learning Deep Structured Multi-Scale Features using Attention-Gated CRFs for Contour Prediction. Advances in Neural Information Processing Systems, Dec 2017, Long Beach, United States. pp.3961-3970. ⟨hal-01646112⟩