Neuroimage May 2023 PMID 36914105

Inter-individual deep image reconstruction via hierarchical neural code conversion.

Jun Kai Ho, Tomoyasu Horikawa, Kei Majima, Fan Cheng, Yukiyasu Kamitani

Full text PDF Download social card

Mentions

Peak: 3 in Mar 2023

across 3 sources since Mar 2023

Extended reach

62,615

potential audience

Total likes

Video views

2,523

Mentions per month

Mar 2023 – Mar 2024

Where it's discussed

X / Twitter5 posts
YouTube2 videos
News2 articles

AI synopsis

What people are saying

The discussions mainly focus on the significance of the methodology, shown by mentions of reducing training data and specific figures like Fig 6 and S13, S14, which demonstrate technical aspects. The tone conveys curiosity and recognition of potential impact, but the depth of dialogue is moderate.

AgreementModerate agreement

Most discussions acknowledge the significance of the research, with some expressing support or approval, such as 'Published' and mentions of data reduction efforts.

InterestHigh interest

The posts show a high level of curiosity, especially with references to specific figures and techniques, indicating genuine interest in the methodology.

EngagementModerate engagement

Participants engage by referencing specific parts of the study like figures and data samples, but the overall tone remains somewhat brief and straightforward.

ImpactModerate impact

The discussions suggest that the research could have a meaningful influence on brain-computer interface development and neural decoding, though this is mostly implied rather than explicitly argued.

Abstract in brief

This study introduces a neural code converter that can translate brain activity patterns between different individuals during visual perception, allowing for comparison and shared understanding of perceptual content.
The results show that hierarchical and detailed visual information are preserved through the conversion process, enabling accurate reconstruction of visual images from brain activity across different people.
Importantly, the method works well even with limited training data and benefits from pooling data across multiple individuals, suggesting potential for personalized brain-computer interfaces and improved cross-subject neural decoding.

Read the full abstract

All mentions

Inter-individual Deep Image Reconstruction Using Neural Code Conversion

This video examines a neural code converter that translates brain activity patterns between individuals during visual perception, enabling accurate image reconstruction and shared perceptual understanding.

KamitaniLab · Jun 21, 2023 1,511 views

Hierarchical Neural Code Conversion for Cross-Individual Image Reconstruction

This study introduces a neural code converter that translates brain activity patterns between individuals during visual perception, enabling accurate cross-person image reconstruction and shared perceptual understanding.

KamitaniLab · Jun 21, 2023 1,013 views

Yuki Kamitani

@ykamit · X

@humanscotti For our early attempt to reduce training data, see https://t.co/wctrelFg9c (Fig 6, S13, S14; 300 samples ≈ 40 mins).

Mar 20, 2024 · 2 likes · 16,009 followers View post →

（⚆_⚆）

@37dogobu · X

https://t.co/wE1M4hRKqC

Jul 4, 2023 · 69 followers View post →

ごまふあざらし(GomahuAzarashi)

@MathSorcerer · X

RT @ykamit: Published https://t.co/wctrelFg9c

Mar 16, 2023 · 1 share · 2,950 followers View post →

トクロンティヌス

@tokurontinus · X