Rethinking Vision Transformers for MobileNet Size and Speed

TL;DR

This paper introduces EfficientFormerV2, a vision transformer architecture optimized for mobile devices that achieves higher accuracy than MobileNetV2 while maintaining similar size and latency, challenging the notion that transformers are inherently less efficient.

Contribution

It proposes a novel supernet and a fine-grained joint search strategy to design efficient transformer architectures suitable for resource-constrained environments.

Findings

EfficientFormerV2 outperforms MobileNetV2 by 3.5% top-1 accuracy on ImageNet-1K.

The models achieve similar latency and parameter count as MobileNetV2.

Properly designed transformers can match MobileNet-level efficiency and performance.

Abstract

With the success of Vision Transformers (ViTs) in computer vision tasks, recent arts try to optimize the performance and complexity of ViTs to enable efficient deployment on mobile devices. Multiple approaches are proposed to accelerate attention mechanism, improve inefficient designs, or incorporate mobile-friendly lightweight convolutions to form hybrid architectures. However, ViT and its variants still have higher latency or considerably more parameters than lightweight CNNs, even true for the years-old MobileNet. In practice, latency and size are both crucial for efficient deployment on resource-constraint hardware. In this work, we investigate a central question, can transformer models run as fast as MobileNet and maintain a similar size? We revisit the design choices of ViTs and propose a novel supernet with low latency and high parameter efficiency. We further introduce a novel fine-grained joint search strategy for transformer models that can find efficient architectures by optimizing latency and number of parameters simultaneously. The proposed models, EfficientFormerV2, achieve 3.5% higher top-1 accuracy than MobileNetV2 on ImageNet-1K with similar latency and parameters. This work demonstrate that properly designed and optimized vision transformers can achieve high performance even with MobileNet-level size and speed.