Learning Discriminative Feature with CRF for Unsupervised Video Object Segmentation

Mingmin Zhen, Shiwei Li, Lei Zhou, Jiaxiang Shang, Haoan Feng, Tian Fang, Long Quan ;

In this paper, we introduce a novel network, called discriminative feature network (DFNet), to address the unsupervised video object segmentation task. To capture the inherent correlation among video frames, we learn K discriminative features (D-features) from the input image and reference images that reveal feature distribution from a global perspective. The D-features are then used to establish correspondence with all features of input image under conditional random field (CRF) formulation, which is leveraged to boost consistency between pixels. The experiments verify that DFNet outperforms state-of-the-art methods by a large margin with a mean IoU score of 83.4\% and ranks first on the DAVIS-2016 leaderboard while using much fewer parameters and achieving much faster speed during inference phase. We further evaluate DFNet on the FBMS dataset and the video saliency dataset ViSal, reaching a new state-of-the-art. To further demonstrate the generalizability of our framework, DFNet is also applied to the image object co-segmentation task. We perform experiments on a challenging dataset PASCAL-VOC and observe the superiority of DFNet. The thorough experiments verify that DFNet is able to capture and mine the underlying relations of images and discover the common foreground objects. "