ActiveKD combines active learning with VLM teacher distillation; PCoreSet selects unlabeled data, delivering up to 29.07% gain on ImageNet and leading in 64 of 73 test settings. Read more: https://getnews.me/activekd-and-pcoreset-boost-active-learning-via-knowledge-distillation/ #activekd #pcore
October 3, 2025 at 8:12 AM
ActiveKD combines active learning with VLM teacher distillation; PCoreSet selects unlabeled data, delivering up to 29.07% gain on ImageNet and leading in 64 of 73 test settings. Read more: https://getnews.me/activekd-and-pcoreset-boost-active-learning-via-knowledge-distillation/ #activekd #pcore
transfer of teacher knowledge under limited annotation budgets. Evaluations on 11 datasets show that PCoreSet consistently outperforms existing selection methods within the ActiveKD framework, advancing research at the intersection of AL and KD. [6/6 of https://arxiv.org/abs/2506.00910v1]
June 3, 2025 at 6:24 AM
transfer of teacher knowledge under limited annotation budgets. Evaluations on 11 datasets show that PCoreSet consistently outperforms existing selection methods within the ActiveKD framework, advancing research at the intersection of AL and KD. [6/6 of https://arxiv.org/abs/2506.00910v1]
teacher models are often unavailable. In this paper, we introduce ActiveKD, a framework that integrates AL with KD by leveraging the zero- and few-shot capabilities of large vision-language models (VLMs). A key aspect of ActiveKD is the structured [3/6 of https://arxiv.org/abs/2506.00910v1]
June 3, 2025 at 6:24 AM
teacher models are often unavailable. In this paper, we introduce ActiveKD, a framework that integrates AL with KD by leveraging the zero- and few-shot capabilities of large vision-language models (VLMs). A key aspect of ActiveKD is the structured [3/6 of https://arxiv.org/abs/2506.00910v1]