Deep Label Prior: Pre-Training-Free Salient Object Detection Network based on Label Learning

• Published in: IEEE transactions on multimedia 2023-01, Vol.25, p.1-13
• Main Authors: Yue, Huihui, Guo, Jichang, Yin, Xiangjun, Zhang, Yi, Zheng, Sida
• Format: Article
• Language: English
• Subjects: Datasets; deep image prior; deep label prior; Deep learning; Electronics packaging; Feature extraction; Image processing; Labels; Morphology; morphology operation; Object detection; Object recognition; Salience; Salient object detection; Semantics; Task analysis; Teaching methods; Training
• ISSN: 1520-9210; 1941-0077
• DOI: 10.1109/TMM.2022.3204440

Due to the excellent semantics extraction capabilities, deep learning methods have significantly progressed in salient object detection (SOD). However, these methods often require time-consuming pre-training and large training datasets with ground truth. To address these issues, by referring to the framework known as "deep image prior (DIP)," we propose a SOD method called deep label prior network (DLPNet), which consists of \cal A-stream and \cal B-stream. The \cal A-stream includes two cascaded UNets and a simple CNNs module to extract the initial saliency map, while the \cal B-stream contains only two cascaded UNets, which refines the extracted initial saliency map. Unlike most of the current deep learning methods, DLPNet views the SOD task as a conditional image generation problem, relying on only the internal prior of the input itself to generate the saliency map. Hence, our DLPNet does not require pre-training or large annotated / unannotated datasets. Furthermore, we propose a morphology operation scheme, which creates rich pseudo-labels for facilitating the updating of network weights. Extensive experiments demonstrate that our method outperforms state-of-the-art unsupervised techniques and is even comparable to state-of-the-art supervised and weakly supervised methods on different evaluation metrics.