Learning Non-target Knowledge for Few-shot Semantic Segmentation

Existing studies in few-shot semantic segmentation only focus on mining the target object information, however, often are hard to tell ambiguous regions, especially in nontarget regions, which include background (BG) and Distracting Objects (DOs). To alleviate this problem, we propose a novel framework, namely Non-Target Region Eliminating (NTRE) network, to explicitly mine and eliminate BG and DO regions in the query. First, a BG Mining Module (BGMM) is proposed to extract the BG region via learning a general BG prototype. To this end, we design a BG loss to supervise the learning of BGMM only using the known target object segmentation ground truth. Then, a BG Eliminating Module and a DO Eliminating Module are proposed to successively filter out the BG and DO information from the query feature, based on which we can obtain a BG and DO free target object segmentation result. Furthermore, we propose a prototypical contrastive learning algorithm to improve the model ability of distinguishing the target object from DOs. Extensive experiments on both PASCAL-5(i) and COCO-20(i) datasets show that our approach is effective despite its simplicity. Code is available at https://github.com/LIUYANWEI98/NERTNet

Automated summary

This study proposes a Non-Target Region Eliminating (NTRE) network framework to explicitly mine and eliminate background and distracting object regions in the query. The proposed framework effectively distinguishes the target object from distracting objects, and experimental results demonstrate its effectiveness.