UniFS: Universal Few-shot Instance Perception with Point Representations

arXiv:2404.19401v1 Announce Type: new
Abstract: Instance perception tasks (object detection, instance segmentation, pose estimation, counting) play a key role in industrial applications of visual models. As supervised learning methods suffer from high labeling cost, few-shot learning methods which effectively learn from a limited number of labeled examples are desired. Existing few-shot learning methods primarily focus on a restricted set of tasks, presumably due to the challenges involved in designing a generic model capable of representing diverse tasks in a unified …

abstract applications arxiv cost cs.cv detection examples few-shot few-shot learning industrial instance key labeling learn object perception role segmentation supervised learning tasks type universal visual