SAM-Lightening: A Lightweight Segment Anything Model with Dilated Flash Attention to Achieve 30 times Acceleration

TL;DR

SAM-Lightening introduces a re-engineered attention mechanism that significantly accelerates segmentation inference speed while reducing memory usage, making SAM more practical for real-world applications.

Contribution

It proposes Dilated Flash Attention and a progressive distillation method to enhance SAM's efficiency without retraining from scratch.

Findings

Achieves 7 ms inference per image of size 1024x1024

30.1 times faster than vanilla SAM

Uses only 244MB memory, 3.5% of vanilla SAM

Abstract

Segment Anything Model (SAM) has garnered significant attention in segmentation tasks due to their zero-shot generalization ability. However, a broader application of SAMs to real-world practice has been restricted by their low inference speed and high computational memory demands, which mainly stem from the attention mechanism. Existing work concentrated on optimizing the encoder, yet has not adequately addressed the inefficiency of the attention mechanism itself, even when distilled to a smaller model, which thus leaves space for further improvement. In response, we introduce SAM-Lightening, a variant of SAM, that features a re-engineered attention mechanism, termed Dilated Flash Attention. It not only facilitates higher parallelism, enhancing processing efficiency but also retains compatibility with the existing FlashAttention. Correspondingly, we propose a progressive distillation to enable an efficient knowledge transfer from the vanilla SAM without costly training from scratch. Experiments on COCO and LVIS reveal that SAM-Lightening significantly outperforms the state-of-the-art methods in both run-time efficiency and segmentation accuracy. Specifically, it can achieve an inference speed of 7 milliseconds (ms) per image, for images of size 1024*1024 pixels, which is 30.1 times faster than the vanilla SAM and 2.1 times than the state-of-the-art. Moreover, it takes only 244MB memory, which is 3.5\% of the vanilla SAM. The code and weights are available at https://anonymous.4open.science/r/SAM-LIGHTENING-BC25/.