# Searching the Search Space of Vision Transformer

Source: [Searching the Search Space of Vision Transformer](<https://www.sota2.com/research/paper/68fa6b4c1e209131cea2d64e>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-01-04T05:12:49.038Z

Authors: Minghao Chen, Kan Wu, Bolin Ni, Houwen Peng, Bei Liu, Jianlong Fu, Hongyang Chao, Haibin Ling

Published: 2021-11-29T17:26:07.000Z

## Abstract

Vision Transformer has shown great visual representation power in substantial vision tasks such as recognition and detection, and thus been attracting fast-growing efforts on manually designing more effective architectures. In this paper, we propose to use neural architecture search to automate this process, by searching not only the architecture but also the search space. The central idea is to gradually evolve different search dimensions guided by their E-T Error computed using a weight-sharing supernet. Moreover, we provide design guidelines of general vision transformers with extensive analysis according to the space searching process, which could promote the understanding of vision transformer. Remarkably, the searched models, named S3 (short for Searching the Search Space), from the searched space achieve superior performance to recently proposed models, such as Swin, DeiT and ViT, when evaluated on ImageNet. The effectiveness of S3 is also illustrated on object detection, semantic segmentation and visual question answering, demonstrating its generality to downstream vision and vision-language tasks. Code and models will be available at https://github.com/microsoft/Cream.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Instance Segmentation on COCO 2017 (val)](<https://www.sota2.com/research/sota/instance-segmentation-on-coco-2017-val.md>): Instance Segmentation; COCO 2017 (val); APm: 0.455.

- [Visual Question Answering on VQA 2.0 (test-dev)](<https://www.sota2.com/research/sota/visual-question-answering-on-vqa-2-0-test-dev.md>): Visual Question Answering; VQA 2.0 (test-dev); Accuracy: 67.84.

- [Visual Question Answering on VQA v2 (test-std)](<https://www.sota2.com/research/sota/visual-question-answering-on-vqa-v2-test-std.md>): Visual Question Answering; VQA v2 (test-std); Accuracy: 68.

- [Semantic Segmentation on ADE20K (val)](<https://www.sota2.com/research/sota/semantic-segmentation-on-ade20k-val.md>): Semantic segmentation; ADE20K (val); mIoU: 48.04.

- [Object Detection on COCO 2017 (val)](<https://www.sota2.com/research/sota/object-detection-on-coco-2017-val.md>): Object Detection; COCO 2017 (val); AP: not recorded.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2111.14725>)

[PDF](<https://arxiv.org/pdf/2111.14725.pdf>)
