Make It Up: Fake Images, Real Gains in Generalized Few-shot Semantic Segmentation

TL;DR

This paper introduces Syn4Seg, a novel framework that leverages synthetic images and advanced pseudo-label refinement to significantly improve generalized few-shot semantic segmentation performance.

Contribution

Syn4Seg enhances GFSS by generating diverse synthetic images and refining pseudo-labels through a support-guided, multi-stage process, addressing coverage and supervision issues.

Findings

Consistent improvements on PASCAL-5i and COCO-20i datasets.

Effective synthetic data generation increases novel-class coverage.

Refined pseudo-labels lead to more accurate segmentation boundaries.

Abstract

Generalized few-shot semantic segmentation (GFSS) is fundamentally limited by the coverage of novel-class appearances under scarce annotations. While diffusion models can synthesize novel-class images at scale, practical gains are often hindered by insufficient coverage and noisy supervision when masks are unavailable or unreliable. We propose Syn4Seg, a generation-enhanced GFSS framework designed to expand novel-class coverage while improving pseudo-label quality. Syn4Seg first maximizes prompt-space coverage by constructing an embedding-deduplicated prompt bank for each novel class, yielding diverse yet class-consistent synthetic images. It then performs support-guided pseudo-label estimation via a two-stage refinement that i) filters low-consistency regions to obtain high-precision seeds and ii) relabels uncertain pixels with image-adaptive prototypes that combine global (support) and local (image) statistics. Finally, we refine only boundary-band and unlabeled pixels using a constrained SAM-based update to improve contour fidelity without overwriting high-confidence interiors. Extensive experiments on PASCAL-$5^i$ and COCO-$20^i$ demonstrate consistent improvements in both 1-shot and 5-shot settings, highlighting synthetic data as a scalable path for GFSS with reliable masks and precise boundaries.