Video saliency prediction using enhanced spatiotemporal alignment network. (January 2021)

• Record Type: Journal Article
• Authors: Chen, Jin; Song, Huihui; Zhang, Kaihua; Liu, Bo; Liu, Qingshan
• Is Part Of: Pattern recognition. Volume 109(2021)
• Publication Date: 2021-01
• Subjects: Video saliency prediction -- Feature alignment -- Deformable convolution -- Bidirectional ConvLSTM; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2020.107615
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: A novel MDAN module is designed to align the features across frames with deformable convolution (This is the first work to apply Dconv to video saliency prediction (VSP). A novel Bi ConvLSTM is introduced to VSP, which makes full use of the long term temporal context information in the forward and backward timing directions. Extensive evaluations on four VSP benchmarks demonstrate the proposed method achieves competing performance against state of the arts. Abstract: Due to a variety of motions across different frames, it is highly challenging to learn an effective spatiotemporal representation for accurate video saliency prediction (VSP). To address this issue, we develop an effective spatiotemporal feature alignment network tailored to VSP, mainly including two key sub-networks: a multi-scale deformable convolutional alignment network (MDAN) and a bidirectional convolutional Long Short-Term Memory (Bi-ConvLSTM) network. The MDAN learns to align the features of the neighboring frames to the reference one in a coarse-to-fine manner, which can well handle various motions. Specifically, the MDAN owns a pyramidal feature hierarchy structure that first leverages deformable convolution (Dconv) to align the lower-resolution features across frames, and then aggregates the aligned features to align the higher-resolution features, progressively enhancing the features from top to bottom. The output of MDAN is then fed into the Bi-ConvLSTM for further enhancement, which captures the useful long-time temporal information along forward and backward timing directions to effectively guide attention orientation shift prediction under complex scene transformation. Finally, the enhanced features are decoded to generate the predicted saliency map. The proposed model is trained end-to-end without any intricate post processing. Extensive evaluations on four VSP benchmark datasets demonstrate that the proposed method achieves favorable performance against state-of-the-art methods. The source codes and all the results will be released at https://github.com/cj4L/ESAN-VSP .