SAILS: Segment Anything with Incrementally Learned Semantics for Task-Invariant and Training-Free Continual Learning

TL;DR

SAILS introduces a training-free, task-invariant continual learning framework for semantic segmentation that leverages foundational models and prototypes, eliminating forgetting and outperforming training-based methods on standard datasets.

Contribution

SAILS proposes a novel training-free approach for class-incremental semantic segmentation using foundational models and prototypes, avoiding retraining and reducing forgetting.

Findings

SAILS outperforms existing training-based methods on standard datasets.

It completely eliminates forgetting by avoiding parameter updates.

SAILS shows positive backward transfer, improving previous class performance with new classes.

Abstract

Continual learning remains constrained by the need for repeated retraining, high computational costs, and the persistent challenge of forgetting. These factors significantly limit the applicability of continuous learning in real-world settings, as iterative model updates require significant computational resources and inherently exacerbate forgetting. We present SAILS -- Segment Anything with Incrementally Learned Semantics, a training-free framework for Class-Incremental Semantic Segmentation (CISS) that sidesteps these challenges entirely. SAILS leverages foundational models to decouple CISS into two stages: Zero-shot region extraction using Segment Anything Model (SAM), followed by semantic association through prototypes in a fixed feature space. SAILS incorporates selective intra-class clustering, resulting in multiple prototypes per class to better model intra-class variability. Our results demonstrate that, despite requiring no incremental training, SAILS typically surpasses the performance of existing training-based approaches on standard CISS datasets, particularly in long and challenging task sequences where forgetting tends to be most severe. By avoiding parameter updates, SAILS completely eliminates forgetting and maintains consistent, task-invariant performance. Furthermore, SAILS exhibits positive backward transfer, where the introduction of new classes can enhance performance on previous classes.