Homological invariants of edge ideals of the multiple extended complete split-like graphs
This work proposes a novel representation learning framework based on adaptive multi-scale fusion and contrastive learning to address the limited representational capacity of existing methods in complex scenes. By dynamically integrating multi-granularity features and incorporating a structure-aware contrastive loss, the proposed approach effectively enhances the model’s ability to capture fine-grained semantics and contextual relationships. Extensive experiments demonstrate that the framework consistently outperforms state-of-the-art methods across multiple benchmark datasets, achieving substantial improvements in both accuracy and robustness. These results underscore its potential as a new technical pathway for tackling challenging visual understanding tasks.