Discriminative Consensus Mining with A Thousand Group for More Accurate Co-Salient Object Detection

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
Perustieteiden korkeakoulu | Master's thesis
Visual Computing and Communication
Degree programme
Master's Programme in ICT Innovation
51 + 0
Co-Salient Object Detection (CoSOD) is a rapidly growing task, extended from Salient Object Detection (CoSOD) and Common Object Segmentation (Co-Segmentation). It is aimed at detecting the co-occurring salient object in the given image group. Many effective approaches have been proposed on the basis of existing datasets. However, there is still no standard and efficient training set in CoSOD, which makes it chaotic to choose training sets in the recently proposed CoSOD methods. First, the drawbacks of existing training sets in CoSOD are analyzed in a comprehensive way, and potential improvements are provided to solve existing problems to some extent. In particular, in this thesis, a new CoSOD training set is introduced, named Co-Saliency of ImageNet (CoSINe) dataset. The proposed CoSINe consists of 22,978 images, which are divided into 919 groups, which is the largest number of groups among all existing CoSOD datasets. The images obtained here span a wide variety in terms of categories, surrounding objects, backgrounds, object locations, object sizes, etc. In experiments, models trained on CoSINe can achieve significantly better performance with fewer images compared to all existing datasets, which represents its great effectiveness and quality. Second, to make the most of the proposed CoSINe, a novel CoSOD approach named Hierarchical Instance-aware COnsensus MinEr ( HICOME) is proposed, which efficiently mines the consensus feature from different feature levels and discriminates objects of different classes in an object-aware contrastive way. As extensive experiments show, the proposed HICOME achieves state-of-the-art performance on all the existing CoSOD test sets. Several useful training tricks suitable for training CoSOD models are also provided. Third, practical applications are given using the CoSOD technique to show the effectiveness. Finally, the remaining challenges and potential improvements of CoSOD are discussed to inspire related work in the future. The source code, the dataset, and the online demo will be publicly available at https://github.com/ZhengPeng7/CoSINe.
Laaksonen, Jorma
Thesis advisor
Laaksonen, Jorma
co-salient object detection, CoSOD dataset, metric learning, consensus mining
Other note