HyCTAS: Multi-Objective Hybrid Convolution-Transformer Architecture Search for Real-Time Image Segmentation

TL;DR

HyCTAS introduces a multi-objective architecture search method to efficiently design hybrid convolution-transformer models for real-time image segmentation, balancing accuracy and latency without extensive manual tuning.

Contribution

The paper proposes a novel multi-target supernet search algorithm that effectively integrates multi-head self-attention into high-resolution CNNs for segmentation.

Findings

Achieves competitive real-time segmentation accuracy on Cityscapes, ADE20K, and COCO.

Finds architectures on the Pareto front balancing latency and mIoU.

Demonstrates effectiveness without ImageNet pretraining.

Abstract

Real-time image segmentation demands architectures that preserve fine spatial detail while capturing global context under tight latency and memory budgets. Image segmentation is one of the most fundamental problems in computer vision and has drawn a lot of attention due to its vast applications in image understanding and autonomous driving. However, designing effective and efficient segmentation neural architectures is a labor-intensive process that may require numerous trials by human experts. In this paper, we address the challenge of integrating multi-head self-attention into high-resolution representation CNNs efficiently by leveraging architecture search. Manually replacing convolution layers with multi-head self-attention is non-trivial due to the costly overhead in memory to maintain high resolution. By contrast, we develop a multi-target multi-branch supernet method, which not only fully utilizes the advantages of high-resolution features but also finds the proper location for placing the multi-head self-attention module. Our search algorithm is optimized towards multiple objectives (e.g., latency and mIoU) and is capable of finding architectures on the approximate Pareto front with an arbitrary number of branches in a single search. We further present a series of models via the Hybrid Convolutional-Transformer Architecture Search (HyCTAS) method that searches for the best hybrid combination of lightweight convolution layers and memory-efficient self-attention layers between branches from different resolutions and fuses them at high resolution for both efficiency and effectiveness. On Cityscapes, ADE20K, and COCO, HyCTAS discovers competitive real-time models without ImageNet pretraining, delivering strong accuracy and latency trade-offs. Code and models are available at https://github.com/MarvinYu1995/HyCTAS.