Bootstrapping SparseFormers from Vision Foundation Models

arXiv:2312.01987v2 Announce Type: replace
Abstract: The recently proposed SparseFormer architecture provides an alternative approach to visual understanding by utilizing a significantly lower number of visual tokens via adjusting RoIs, greatly reducing computational costs while still achieving promising performance. However, training SparseFormers from scratch is still expensive, and scaling up the number of parameters can be challenging. In this paper, we propose to bootstrap SparseFormers from ViT-based vision foundation models in a simple and efficient way. Since the majority of SparseFormer …

arxiv bootstrapping cs.cv foundation type vision