S^4M: Boosting Semi-Supervised Instance Segmentation with SAM
This addresses performance constraints in instance segmentation for computer vision applications where labeled data is scarce, representing an incremental improvement over existing teacher-student approaches.
The paper tackles the problem of semi-supervised instance segmentation with limited labeled data by integrating the Segment Anything Model (SAM) into a teacher-student framework, achieving state-of-the-art performance through a novel distillation method and pseudo-label refinement.
Semi-supervised instance segmentation poses challenges due to limited labeled data, causing difficulties in accurately localizing distinct object instances. Current teacher-student frameworks still suffer from performance constraints due to unreliable pseudo-label quality stemming from limited labeled data. While the Segment Anything Model (SAM) offers robust segmentation capabilities at various granularities, directly applying SAM to this task introduces challenges such as class-agnostic predictions and potential over-segmentation. To address these complexities, we carefully integrate SAM into the semi-supervised instance segmentation framework, developing a novel distillation method that effectively captures the precise localization capabilities of SAM without compromising semantic recognition. Furthermore, we incorporate pseudo-label refinement as well as a specialized data augmentation with the refined pseudo-labels, resulting in superior performance. We establish state-of-the-art performance, and provide comprehensive experiments and ablation studies to validate the effectiveness of our proposed approach.