MobileVOS: Real-Time Video Object Segmentation Contrastive Learning meets Knowledge Distillation

This paper tackles the problem of semi-supervised video object segmentation on resource-constrained devices, such as mobile phones. We formulate this problem as a distillation task, whereby we demonstrate that small spacetime-memory networks with finite memory can achieve competitive results with state of the art, but at a fraction of the computational cost (32 milliseconds per frame on a Samsung Galaxy S22). Specifically, we provide a theoretically grounded framework that unifies knowledge distillation with supervised contrastive representation learning. These models are able to jointly benefit from both pixel-wise contrastive learning and distillation from a pretrained teacher. We validate this loss by achieving competitive J &F to state of the art on both the standard DAVIS and YouTube benchmarks, despite running up to ˆ5 faster, and with ˆ32 fewer parameters.