Skip to main navigation Skip to search Skip to main content

Cross-Subject Mind Decoding from Inaccurate Representations

  • Yangyang Xu*
  • , Bangzhen Liu
  • , Wenqi Shao
  • , Yong Du*
  • , Shengfeng He
  • , Tingting Zhu
  • *Corresponding author for this work
  • University of Oxford
  • South China University of Technology
  • Singapore Management University
  • Shanghai Artificial Intelligence Laboratory
  • Ocean University of China

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Decoding stimulus images from fMRI signals has advanced with pre-trained generative models. However, existing methods struggle with cross-subject mappings due to cognitive variability and subject-specific differences. This challenge arises from sequential errors, where unidirectional mappings generate partially inaccurate representations that, when fed into diffusion models, accumulate errors and degrade reconstruction fidelity. To address this, we propose the Bidirectional Autoencoder Intertwining framework for accurate decoded representation prediction. Our approach unifies multiple subjects through a Subject Bias Modulation Module while leveraging bidirectional mapping to better capture data distributions for precise representation prediction. To further enhance fidelity when decoding representations into stimulus images, we introduce a Semantic Refinement Module to improve semantic representations and a Visual Coherence Module to mitigate the effects of inaccurate visual representations. Integrated with ControlNet and Stable Diffusion, our method outperforms state-of-the-art approaches on benchmark datasets in both qualitative and quantitative evaluations. Moreover, our framework exhibits strong adaptability to new subjects with minimal training samples.

Original languageEnglish
Title of host publicationProceedings - 2025 IEEE/CVF International Conference on Computer Vision, ICCV 2025
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages15066-15075
Number of pages10
ISBN (Electronic)9798331587758
DOIs
StatePublished - 2025
Externally publishedYes
Event2025 IEEE/CVF International Conference on Computer Vision, ICCV 2025 - Honolulu, United States
Duration: 19 Oct 202523 Oct 2025

Publication series

NameProceedings of the IEEE International Conference on Computer Vision
ISSN (Print)1550-5499
ISSN (Electronic)2380-7504

Conference

Conference2025 IEEE/CVF International Conference on Computer Vision, ICCV 2025
Country/TerritoryUnited States
CityHonolulu
Period19/10/2523/10/25

Fingerprint

Dive into the research topics of 'Cross-Subject Mind Decoding from Inaccurate Representations'. Together they form a unique fingerprint.

Cite this