Preserving Cross-Modal Consistency for CLIP-based Class-Incremental Learning
PositiveArtificial Intelligence
- The paper presents a novel framework, DMC, designed to improve class
- The development of DMC is significant as it addresses the critical issue of knowledge retention in AI models, allowing them to adapt to new tasks while preserving previously learned information, which is essential for advancing AI capabilities.
- While there are no directly related articles, the focus on datasets like CIFAR
— via World Pulse Now AI Editorial System