Pith. sign in

REVIEW 2 cited by

Improving Vision Transformers by Revisiting High-frequency Components

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2204.00993 v3 pith:YFTXXA2E submitted 2022-04-03 cs.CV

Improving Vision Transformers by Revisiting High-frequency Components

classification cs.CV
keywords modelscomponentshigh-frequencytrainingvisiondatafrequencyimages
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

The transformer models have shown promising effectiveness in dealing with various vision tasks. However, compared with training Convolutional Neural Network (CNN) models, training Vision Transformer (ViT) models is more difficult and relies on the large-scale training set. To explain this observation we make a hypothesis that \textit{ViT models are less effective in capturing the high-frequency components of images than CNN models}, and verify it by a frequency analysis. Inspired by this finding, we first investigate the effects of existing techniques for improving ViT models from a new frequency perspective, and find that the success of some techniques (e.g., RandAugment) can be attributed to the better usage of the high-frequency components. Then, to compensate for this insufficient ability of ViT models, we propose HAT, which directly augments high-frequency components of images via adversarial training. We show that HAT can consistently boost the performance of various ViT models (e.g., +1.2% for ViT-B, +0.5% for Swin-B), and especially enhance the advanced model VOLO-D5 to 87.3% that only uses ImageNet-1K data, and the superiority can also be maintained on out-of-distribution data and transferred to downstream tasks. The code is available at: https://github.com/jiawangbai/HAT.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. OmniDS: Dual-Stream Context Fusion for Omnidirectional Depth from Fisheye Cameras

    cs.CV 2026-07 conditional novelty 6.0

    OmniDS reaches SOTA inverse-depth accuracy on OmniThings/OmniHouse/Sunny by iterative dual-stream ERP context fusion plus consensus volumes, with a distilled MobileNet variant at ~10 FPS.

  2. Scale-Aware Adversarial Analysis: A Diagnostic for Generative AI in Multiscale Complex Systems

    cs.LG 2026-05 unverdicted novelty 6.0

    A new scale-aware diagnostic framework shows that unconstrained diffusion generative models exhibit structural freezing and instability instead of smooth physical responses under multiscale perturbations.