Weakly Supervised From survey PDF — verify links
DAKD
DAKD — Knowledge Distillation
WACV · 2025Human-reviewed summary status
Survey summary
Distills knowledge from aggregated representations of multiple backbones into a single-backbone student model via bi-level distillation and a disentangled cross-attention feature-aggregation network, improving over prior methods on UCF-Crime, ShanghaiTech, and XD-Violence.
Taxonomy placement
Weakly Supervised
Weakly SupervisedKnowledge Distillation
Method characteristics
Weakly SupervisedAggregated knowledgeDistillation