EdgeTAM: On-Device Track Anything Model

TL;DR

EdgeTAM significantly enhances the efficiency of the SAM 2 video segmentation model, enabling real-time on-device processing on mobile devices while maintaining high accuracy through a novel 2D Spatial Perceiver and distillation techniques.

Contribution

The paper introduces EdgeTAM, a novel approach that reduces computational cost of SAM 2 using a 2D Spatial Perceiver and a distillation pipeline, enabling mobile device deployment.

Findings

Achieves high segmentation accuracy on multiple benchmarks.

Runs at 16 FPS on iPhone 15 Pro Max.

Maintains performance comparable to state-of-the-art models.

Abstract

On top of Segment Anything Model (SAM), SAM 2 further extends its capability from image to video inputs through a memory bank mechanism and obtains a remarkable performance compared with previous methods, making it a foundation model for video segmentation task. In this paper, we aim at making SAM 2 much more efficient so that it even runs on mobile devices while maintaining a comparable performance. Despite several works optimizing SAM for better efficiency, we find they are not sufficient for SAM 2 because they all focus on compressing the image encoder, while our benchmark shows that the newly introduced memory attention blocks are also the latency bottleneck. Given this observation, we propose EdgeTAM, which leverages a novel 2D Spatial Perceiver to reduce the computational cost. In particular, the proposed 2D Spatial Perceiver encodes the densely stored frame-level memories with a lightweight Transformer that contains a fixed set of learnable queries. Given that video segmentation is a dense prediction task, we find preserving the spatial structure of the memories is essential so that the queries are split into global-level and patch-level groups. We also propose a distillation pipeline that further improves the performance without inference overhead. As a result, EdgeTAM achieves 87.7, 70.0, 72.3, and 71.7 J&F on DAVIS 2017, MOSE, SA-V val, and SA-V test, while running at 16 FPS on iPhone 15 Pro Max.