Distill-then-prune: An Efficient Compression Framework for Real-time Stereo Matching Network on Edge Devices

arXiv:2405.11809v1 Announce Type: new
Abstract: In recent years, numerous real-time stereo matching methods have been introduced, but they often lack accuracy. These methods attempt to improve accuracy by introducing new modules or integrating traditional methods. However, the improvements are only modest. In this paper, we propose a novel strategy by incorporating knowledge distillation and model pruning to overcome the inherent trade-off between speed and accuracy. As a result, we obtained a model that maintains real-time performance while delivering high accuracy …

abstract accuracy arxiv compression cs.ai cs.cv devices edge edge devices framework however improvements modules network paper real-time type