Continual Learning
To investigate SAM's potential in the continual scenario, we construct a benchmark for continual segmentation, called Continual SAM Adaptation Benchmark (CoSAM), which aims to systematically evaluate SAM-related algorithms's performance in the streaming scenarios. Specifically, CoSAM offers a set of 8 tasks covering diverse domains, including industrial defects, medical imaging, and camouflaged objects, to serve as a realistic and effective benchmark for evaluating current methods.
- Categories:
18 Views
Concept-1K is a novel dataset designed to facilitate research on incremental learning in large language models. It comprises 1,023 concepts represented as knowledge triplets, focusing on recently emerged topics to minimize data leakage. By providing a fine-grained approach to evaluating model performance, Concept-1K enhances the understanding of how these models learn and retain new information.
- Categories:
29 Views