Projection-Free CLIP-Scale EEG Latents via a U-Net-Style Autoencoder.

Overview

1. School of Computer Science and Engineering, Soongsil University, 369 Sangdo-ro, Dongjak-gu, Seoul 06978, Republic of Korea
2. Sweetndata Inc. Seoul 06224, Republic of Korea; (J.A.); (J.S.)
3. Department of Digital Media Engineering, The Catholic University of Korea, Bucheon-si 14662, Republic of Korea

Abstract

Electroencephalography is emerging as a promising conditioning modality for generative visual models. However, existing representation learning approaches often rely on high-capacity masked autoencoders and complex projection networks. When constrained to compact embedding dimensions to match vision-language models, these heavy transformer-based bottlenecks frequently suffer from representation collapse and lose critical signal dynamics. To address this, we propose a lightweight and projection-free autoencoder that directly outputs compact, Contrastive Language–Image Pre-training (CLIP)-scale latent vectors trained toward the CLIP embedding space. Our model adopts a U-Net-style architecture combining one-dimensional convolutional residual blocks for temporal dynamics and inter-channel attention modules for spatial dependencies, alongside skip connections to ensure stable reconstruction. Extensive experiments on visual perception datasets demonstrate that our approach successfully tracks complex signal amplitudes without collapsing. Under strict dimensional constraints, the proposed model achieves superior signal reconstruction fidelity across time and frequency domains using significantly fewer parameters than traditional masked autoencoder baselines. Furthermore, latent space visualizations and zero-shot retrieval tasks reveal that while the baseline collapses toward unstructured, near-chance representations, our architecture preserves emerging, partial semantic organization and retrieves several times above chance. This indicates that the proposed design preserves signal structure while exhibiting preliminary, above-chance semantic alignment, enabling integration into brain-driven generative pipelines.

Reproduced under the paper's license (CC BY), from the paper cited above.

Code

The paper links to its data, not to its authors' code: see the Data section.

The paper's code and data availability statement is in the Data section.

Tracing map

A tracing map links a paper to the code its authors published: this paper has none, so it has no map.

Data

Datasets cited

• github.com/perceivelab/eeg_visual_classification, at github.com; found in “Data Availability Statement”

Data Availability Statement

The EEGCVPR40 and THINGS EEG2 datasets used in this study are publicly available from their respective original publications [37,38,39]. The EEGCVPR40 dataset is available at https://github.com/perceivelab/eeg_visual_classification (accessed on 15 July 2026).

Reproduced under the paper's license (CC BY), from the paper cited above.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 27 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 4 authors, 6 keywords, 1 funder, 21 references.

Cite

This paper

Lee, J., Ahn, J., Sim, J., & Kang, H. (2026). Projection-Free CLIP-Scale EEG Latents via a U-Net-Style Autoencoder. Sensors (Basel, Switzerland), 26(14), 4583. https://doi.org/10.3390/s26144583

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.1038/s41598-026-68186-2 [code]

NeuroStream: spectral-spatio-temporal deep learning for visual stimulus classification from EEG.

Journal: Scientific reports

In common: EEG, 3 references

[2] doi:10.1371/journal.pone.0351872 [code]

Decoding visual object recognition from EEG signals.

Journal: PloS one

In common: EEG, 3 references

[3] doi:10.1038/s41467-026-75653-x [code]

Concept2Brain: an AI model for predicting neurophysiological responses to text and pictures.

Journal: Nature communications

In common: EEG, 2 references

[4] doi:10.1186/s12880-026-02481-2 [code]

Deep learning-based neuroanatomical profiling reveals population-specific brain changes in multiple sclerosis: a large-scale Middle Eastern study.

Journal: BMC medical imaging

In common: 2 references

[5] doi:10.3389/fnins.2026.1810609

A dual-branch network with brain region-constrained attention for EEG emotion recognition.

Journal: Frontiers in neuroscience

In common: EEG, 1 reference

[6] doi:10.3390/s26134045

Brain Signal for Secure EEG Biometric Authentication: A Comprehensive Survey.

Journal: Sensors (Basel, Switzerland)

In common: EEG, 1 reference

[7] doi:10.1002/mpr.70088

Explainable Temporal Deep Learning for EEG-Based Depression Detection Using Resting-State Brain Dynamics.

Journal: International journal of methods in psychiatric research

In common: EEG, 1 reference

[8] doi:10.3389/fnins.2026.1851006

Domain-aware domain-class adaptation network for motor execution to motor imagery EEG classification.

Journal: Frontiers in neuroscience

In common: EEG, 1 reference

[9] doi:10.1007/s10548-026-01238-y [code]

Topographic Reorganization of EEG Complexity During Visual Mental Imagery: Insights from Lempel-Ziv Complexity in High-Density EEG.

Journal: Brain topography

In common: EEG, 1 reference

[10] doi:10.1162/imag.a.1207 [code]

Investigating the temporal dynamics and modeling of mid-level feature representations in humans.

Journal: Imaging neuroscience (Cambridge, Mass.)

In common: EEG, 1 reference

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.