3 accepted papers
Anticipating future actions in videos is challenging, as the observed frames provide only evidence of past activities, requiring the inference of latent intentions to predict upcoming actions. Existing transformer-based approaches, which rely on dot-product attention over pixel representations, ofte…
Training-free Neural Architecture Search (NAS) efficiently identifies high-performing neural networks using zero-cost (ZC) proxies. Unlike multi-shot and one-shot NAS approaches, ZC-NAS is both (i) time-efficient, eliminating the need for model training, and (ii) interpretable, with proxy designs of…
The landscape of deep learning research is moving towards innovative strategies to harness the true potential of data. Traditionally emphasis has been on scaling model architectures resulting in large and complex neural networks which can be difficult to train with limited computational resources. H…