3D-Telepathy: Reconstructing 3D Objects from EEG Signals

Ahmed Elazab; Changmiao Wang; Gangyong Jia; Jionghao Cheng; Nannan Li; Ruiquan Ge; Xiang Wan; Yuxiang Ge; Zhaojie Fang

REVIEW 5 major objections 4 minor 24 references

3D-Telepathy: Reconstructing 3D Objects from EEG Signals

T0 review · 5 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A dual self-attention EEG encoder can condition a NeRF through variational score distillation to generate 3D objects whose content and structure match the original visual stimuli.

desk verdict A plausible first step toward EEG-to-3D, but the evidence is not yet there: no held-out split, no control, and a wrong dataset citation. read the letter →

arxiv 2506.21843 v1 pith:N6PJ3MXB submitted 2025-06-27 cs.CV

Yuxiang Ge , Jionghao Cheng , Ruiquan Ge , Zhaojie Fang , Gangyong Jia , Xiang Wan , Nannan Li , Ahmed Elazab

show 1 more author

Changmiao Wang

This is my paper · ORCID

classification cs.CV

keywords EEGdecoding3Dobjectreconstructionneuralradiancefieldsvariationalscoredistillationstablediffusionpriordualself-attentionbrain-computerinterfacecross-modalgeneration

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper reports reconstructing 3D visual objects directly from EEG brain signals, a step past the usual EEG-to-2D image reconstruction task. The authors reason that the brain encodes spatial structure even when a person views flat images, so a successful decoder should recover 3D form rather than discard depth. They build a pipeline whose EEG encoder uses a dual self-attention mechanism and is trained with masked self-supervision, contrastive learning, and cross-attention, and whose 3D generator uses Stable Diffusion as a prior to train a NeRF through variational score distillation. On frames rendered from the NeRF and aligned to the original stimulus viewpoints, the pipeline reports better PSNR, LPIPS, and FID than several EEG-to-2D baselines evaluated the same way. If the claim holds, EEG contains enough visual-semantic information to condition volumetric generation, which would matter for brain-computer interfaces and assistive communication.

What carries the argument

The load-bearing mechanism is the EEG embedding $\xi^e_i$ extracted by the dual self-attention encoder. The first attention layer operates over electrode positions to form EEG tokens; a second layer, augmented by a record token, folds those tokens into a single condition vector. That vector is passed through the Q/K/V projections of cross-attention into a fine-tuned Stable Diffusion U-Net, and variational score distillation (VSD) updates the NeRF parameters $\theta$ using the difference between the pretrained diffusion score and a LoRA-adapted score for rendered images, as in equation (10). The dual self-attention is what turns noisy multichannel EEG into a stable condition, and VSD is what lets a 2D diffusion prior supervise the 3D radiance field.

What would settle it

Compare the same pipeline against a strict held-out split of subjects or object categories, and score the 3D output with a geometric metric such as Chamfer distance between the reconstructed surface and the ground-truth object, or depth-map accuracy against the rendered stimulus. If PSNR falls to chance-level or the geometric metrics are no better than a generic mean-object prior, the claim that EEG drives genuine 3D reconstruction is not supported.

Watch

Extended reading notes

Core claim

The central claim is that 3D-Telepathy establishes an end-to-end mapping from raw EEG recordings to 3D objects. A dual self-attention encoder first turns multichannel, multi-timepoint EEG into a single condition embedding; that embedding is injected into the cross-attention layers of a fine-tuned Stable Diffusion U-Net, and variational score distillation then optimizes a NeRF so that its rendered views match the distribution of images the diffusion model thinks are plausible for that EEG condition. The paper reports PSNR 28.5203, LPIPS 0.3178, and FID 57.3326 on aligned rendered frames, the best among the EEG-to-2D systems it compares against. The authors take this as evidence that the spatial information in EEG that is lost in 2D-only reconstructions can be recovered in 3D, while acknowledging that sides of an object never visible in the original stimulus are not intelligently filled in.

Load-bearing premise

The central claim rests on the assumption that the EEG recordings and target images are genuinely paired and cleanly separated into training and evaluation, so the encoder must decode visual content rather than memorize the stimulus set; if that pairing or split is unreliable, the reported gains could shrink or disappear under a held-out test.

Editorial extensions

If this is right

The mapping from EEG to 3D is end-to-end: one trained encoder feeds a NeRF, so no manual image-to-3D step is needed after training.
Because the condition vector is injected into cross-attention layers of a diffusion U-Net, the same EEG embedding supports 2D rendering supervision during NeRF optimization.
On viewpoint-aligned frames, the pipeline outperforms existing EEG-to-2D baselines on PSNR, LPIPS, and FID, indicating the 3D output retains image-level fidelity.
The visible sides of an object are reconstructed from the EEG; unseen sides are not filled in, so the method recovers observed structure but does not hallucinate.
A brain-computer interface could present a user's observed stimulus as a navigable 3D object rather than a flat image.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Subject-independent held-out evaluation with true 3D metrics would test whether the reported gains survive generalization; the current 2D-frame protocol cannot by itself distinguish decoding from dataset memorization.
The same encoder-plus-diffusion-prior recipe could be applied to other neural modalities or to video stimuli, where temporal EEG dynamics may encode motion and depth cues not present in still-image training.
Since the ablations show the largest drops when the masked self-supervised and contrastive stage is removed, a simpler contrastive-only encoder might approach the same performance with less training cost.
If EEG truly encodes spatial layout, reconstructing 3D from EEG could serve as a neural probe for how the visual system represents depth and viewpoint; this setup offers a testbed for that question.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

You should know this paper before the hype: it builds a reasonable end-to-end pipeline from EEG to 3D, but the experiments as reported do not establish that the EEG signal, rather than the diffusion prior or in-sample memorization, drives the reconstructions. The central claim is conditional at best.

What's new and worth credit: the task itself is real and underexplored. Prior work stopped at EEG-to-2D; extending to NeRF via variational score distillation is a sensible next step. The dual self-attention EEG encoder and the hybrid training recipe (MAE-style masking, contrastive learning, cross-attention with images) are all known ingredients, but the combination is novel and the ablation study in Table I suggests each piece contributes. They also ship a code link and are candid in the conclusion about texture loss and inability to hallucinate unseen views. That honesty earns goodwill.

The soft spots are serious, though. First, there is no train/test split, no subject-independent evaluation, and no error bars anywhere in Section IV. All reported PSNR/LPIPS/FID numbers appear to be on in-sample reconstructions. If the evaluated EEG trials were used in Stages 1 and 2, those numbers can be inflated by memorization and say nothing about generalization. Second, there is no no-condition or shuffled-EEG baseline. A random or mean embedding fed through the same Stable Diffusion + NeRF prior would likely produce plausible images, so without that control you cannot attribute the output to neural decoding at all. Third, reference [12] is misattributed: MOABB is a BCI benchmarking toolbox, not an EEG-ImageNet dataset. The actual data source, probably an earlier EEG visual decoding dataset, is not properly cited. That is a reproducibility problem. Fourth, judging 3D reconstruction quality by 2D frame metrics after manual viewpoint alignment is clever but indirect; a Chamfer distance or depth comparison against a known 3D mesh would be far more convincing. They do acknowledge the limitation of unseen sides, so this is a known trade-off rather than a hidden flaw.

Who this is for: a reader working on EEG decoding or cross-modal 3D generation will want to see the architecture and the evaluation pitfalls. It deserves a serious referee, not a desk reject, because the task is important and the approach is coherent. But my own verdict is skeptical until the authors add a proper split, a control condition, and fix the dataset citation.

My recommendation: engage with it as a conditional submission. Ask for major revision: report held-out accuracy with subject independence, add a no-condition baseline, and correct the data provenance. If those fixes land, this could become a real contribution.

Referee Report

5 major / 4 minor

Summary. The paper proposes 3D-Telepathy, an end-to-end pipeline that decodes EEG signals into 3D object representations. It introduces a dual self-attention EEG encoder trained with masked autoencoding, contrastive learning, and cross-attention losses, then uses the resulting embeddings as conditioning for a NeRF optimized through variational score distillation (VSD) with a Stable Diffusion prior. Quantitative evaluation compares rendered NeRF frames against static reference images using PSNR, LPIPS, and FID, and the paper reports ablations over training losses and hyperparameters.

Significance. If validated, this would be a first demonstration of 3D visual reconstruction from EEG signals and would extend cross-modal brain decoding into the 3D domain. Strengths include the use of a standard VSD update rule, the combination of self-supervised and contrastive objectives for the EEG encoder, and the candid acknowledgment of limitations such as loss of fine texture and poor fidelity on unseen object sides. The code link is provided, though its contents were not verified during review. However, the current experiments do not establish out-of-sample decoding: no held-out split, subject-independent evaluation, or control condition is reported, and the dataset provenance is questionable. The central claim is therefore not yet adequately supported.

major comments (5)

[Section IV.A, IV.B, IV.D] The evaluation never reports a held-out train/test split, a subject-independent split, or any repeated-run statistics. Section IV.A describes the training data, Section IV.B fine-tunes the U-Net on EEG-image pairs, and Section IV.D reports PSNR/LPIPS/FID on frames aligned with EEG-ImageNet reference images. Because the text does not state that the evaluated stimuli were excluded from Stage 1, Stage 2, and the U-Net fine-tuning, the headline numbers (PSNR 28.5203, LPIPS 0.3178, FID 57.3326) could reflect in-sample memorization rather than generalizable decoding of EEG signals.
[Section III.C and Section IV.D] There is no control condition: the paper does not compare against unconditional generation, a mean-EEG embedding, or shuffled EEG labels. Since the pipeline conditions a strong Stable Diffusion/NeRF prior with CFG=3, such a baseline is necessary to establish that the EEG embedding, rather than the image prior alone, drives the reconstruction content.
[Section IV.A] The dataset citation is not reliable. Reference [12] is the MOABB BCI benchmarking toolbox by Jayaram and Barachant, not an EEG-ImageNet dataset containing approximately 40 categories, 2000 images, and 128-channel EEG recordings during image viewing. The authors must cite the actual dataset and provide details on subjects, number of trials per image, and the exact train/test protocol.
[Section IV.D] The central 3D reconstruction claim is quantified only by 2D frame metrics after a viewpoint-alignment procedure, and no 3D ground-truth metric (e.g., Chamfer distance, depth accuracy, or volume IoU) is reported. The viewpoint-matching step is not described in a way that rules out manual selection of favorable frames, and the paper itself acknowledges that unseen object sides have poor fidelity; this limitation requires quantitative evaluation.
[Section IV.D, Table II] The comparison in Table II is not controlled: CapsEEGNet, Brain2Image, DreamDiffusion, and Dongyang Li et al. are 2D image-reconstruction methods, whereas the proposed method selects 3D-rendered frames for comparison. Differences in PSNR/LPIPS/FID may reflect the evaluation protocol rather than the EEG-decoding capability; results should be compared under the same image-generation setup or at least with the same train/test split.

minor comments (4)

[Section III.C] The abbreviation VSD is expanded as 'Variational Score Diffusion' but the method is Variational Score Distillation; please correct the terminology throughout.
[Section IV.E, Table I] The checkmarks in Table I do not align unambiguously with the four training components, and the percentage changes quoted in the text (23.7%, 123.9%, 228.7%) do not exactly match the table values; please make the table readable and recompute the percentages.
[Section IV.A] The phrase 'padded to maintain 128 channels' is unclear when the acquisition device already used 128 electrodes; specify which channels were missing and how the padding was performed.
[Section IV.D and References [23], [24]] References [23] and [24] are introduced as existing EEG-to-3D works but are not included in the quantitative comparison; please state whether their code or results were unavailable and discuss how this affects the claim of being the first end-to-end EEG-to-3D model.

Circularity Check

1 steps flagged · score 6.0 of 10

The headline PSNR/LPIPS/FID are the selected best row of a hyperparameter sweep computed on the same EEG-ImageNet evaluation images, with no reported held-out split, so the reported 'prediction' is a fitted optimum rather than an out-of-sample result.

fitted input called prediction [Section IV.F and Table III versus Section IV.D and Table II]
""The model reaches optimal performance when γS = 1 and γC = 0.5. Further testing on the re-parameterization of γE and γI reveals that the best results occur when γE = 0.25 and γI = 1.5" (Sec. IV.F); Table III's best row is "1 0.5 0.25 1.5 0.75 0.5 1 28.520 0.318 57.333", and Table II reports "3D-Telepathy (Ours) 28.5203 0.3178 57.3326"."

The numbers presented as the main result are the best configuration found by sweeping γS, γC, γE, γI, EEG/image mask ratios, fine-tuning on/off, CFG, and VSD-versus-SDS, with the same PSNR/LPIPS/FID metrics and the same EEG-ImageNet reference images used for evaluation. The best row of Table III is literally the final row of Table II, so the reported 'prediction' is the selected optimum of the evaluation metric over the evaluation set, not a forecast on unseen EEG trials. Since no train/test or subject split is documented, the comparison to CapsEEGNet, DreamDiffusion, and Dongyang Li inherits this selection and cannot establish generalized EEG-to-3D decoding.

full rationale

The derivation chain itself is not circular: the EEG encoder losses (Eqs. 1-7) and VSD update (Eqs. 8-10) use standard formulation with external image-conditioned training, and the 3D generation stage is a standard VSD/NeRF optimization. No load-bearing result is imported from a same-author citation: the VSD, MAE, LoRA, and diffusion references are independent prior work. The circularity is in the evaluation: the final PSNR/LPIPS/FID are obtained by hyperparameter/model-configuration search on the same metric family and the same EEG-ImageNet images, without a documented held-out split, making the headline numbers a fitted selection rather than a predictive test. Separately, reference [12] cites MOABB, a BCI benchmarking toolbox, not an EEG-ImageNet dataset, so the dataset premise is misattributed; that is a factual/provenance risk rather than a circularity step.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The central claim rests on a dataset assumption (EEG-ImageNet pairing from MOBBA), a signal-sufficiency assumption (EEG carries usable 3D information), the validity of diffusion score distillation as a 3D prior, and a manual evaluation protocol. The tuned loss weights and masking ratios are fitted to the same metrics used for the headline numbers.

free parameters (6)

gamma_C and gamma_S (Stage 1 loss weights) = gamma_C = 0.5, gamma_S = 1
Chosen by sweeping in Table III to maximize PSNR and minimize LPIPS/FID on the evaluation set.
gamma_E and gamma_I (Stage 2 loss weights) = gamma_E = 0.25, gamma_I = 1.5
Selected via the same hyperparameter search in Table III.
EEG mask ratio = 0.75
Masking rate for the masked autoencoder stage, set by experimentation.
Image mask ratio = 0.5
Masking rate for the image autoencoder in Stage 2, tuned via Table III.
Classifier-free guidance (CFG) = 3
The paper says experiments across CFG values showed CFG=3 is best.
VSD particle number = 4
Number of particles for variational score distillation, chosen without reported ablation.

assumptions (5)

domain assumption The MOBBA platform provides a valid EEG-ImageNet paired dataset.
The paper cites reference [12], MOABB, as the source of EEG-ImageNet pairs, but MOABB is a BCI benchmarking suite. If the dataset identity or pairing is wrong, all results lose their grounding.
domain assumption EEG signals contain sufficient 3D spatial information to condition a NeRF.
The entire method assumes the learned EEG embedding retains enough shape and viewpoint information for variational score distillation to produce a coherent 3D object.
standard math Pretrained Stable Diffusion can act as a reliable score oracle for rendered NeRF images.
The VSD derivation in Section III-C inherits the DreamFusion/ProlificDreamer assumption that a 2D diffusion model's score approximates the gradient for 3D optimization.
ad hoc to paper Segmenting only the central object improves 3D generation quality.
Section IV-C states this is based on the finding that including background in 3D video generation gives unsatisfactory results, but no quantitative evidence is provided.
domain assumption Manual viewpoint alignment between NeRF frames and EEG-ImageNet images is accurate enough for metric comparison.
The evaluation in Section IV-D depends on this alignment; the protocol is not specified in enough detail to assess its validity.

how reviews work

0 comments

Cite this review

Pith. "Pith review of 3D-Telepathy: Reconstructing 3D Objects from EEG Signals." pith.science (2026). https://pith.science/paper/N6PJ3MXB

@misc{pith2026250621843,
  author       = {Pith},
  title        = {Pith review of: 3D-Telepathy: Reconstructing 3D Objects from EEG Signals},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/N6PJ3MXB}},
  note         = {Machine review of arXiv:2506.21843}
}

read the original abstract

Reconstructing 3D visual stimuli from Electroencephalography (EEG) data holds significant potential for applications in Brain-Computer Interfaces (BCIs) and aiding individuals with communication disorders. Traditionally, efforts have focused on converting brain activity into 2D images, neglecting the translation of EEG data into 3D objects. This limitation is noteworthy, as the human brain inherently processes three-dimensional spatial information regardless of whether observing 2D images or the real world. The neural activities captured by EEG contain rich spatial information that is inevitably lost when reconstructing only 2D images, thus limiting its practical applications in BCI. The transition from EEG data to 3D object reconstruction faces considerable obstacles. These include the presence of extensive noise within EEG signals and a scarcity of datasets that include both EEG and 3D information, which complicates the extraction process of 3D visual data. Addressing this challenging task, we propose an innovative EEG encoder architecture that integrates a dual self-attention mechanism. We use a hybrid training strategy to train the EEG Encoder, which includes cross-attention, contrastive learning, and self-supervised learning techniques. Additionally, by employing stable diffusion as a prior distribution and utilizing Variational Score Distillation to train a neural radiation field, we successfully generate 3D objects with similar content and structure from EEG data.

Figures

Figures reproduced from arXiv: 2506.21843 by the authors.

**Figure 2.** Overview of the 3D-Telepathy Framework. In Stage 1, we propose an EEG encoder architecture featuring a dual self-attention mechanism. The first [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Our proposed EEG encoder training strategy. In the first step, the [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: The results of 3D reconstruction of an EEG signal using 3D-Telepathy are shown here for [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

**Figure 5.** Figure 5: Different results obtained using VSD or SDS methods and different [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

24 extracted references · 16 canonical work pages

[12]

MOABB: trustworthy algorithm bench- marking for bcis,

V . Jayaram and A. Barachant, “MOABB: trustworthy algorithm bench- marking for bcis,”Journal of Neural Engineering, vol. 15, p. 066011, Sept. 2018

work page 2018
[1]

The human brain in depth: how we see in 3d,

A. E. Welchman, “The human brain in depth: how we see in 3d,”Annual Review of Vision Science, vol. 2, no. 1, pp. 345–376, 2016

work page 2016
[2]

Dreamdiffusion: Generating high-quality images from brain eeg signals,

Y . Bai, X. Wang, Y . Cao, Y . Ge, C. Yuan, and Y . Shan, “Dreamdiffusion: Generating high-quality images from brain eeg signals,”arXiv preprint arXiv:2306.16934, 2023

arXiv 2023
[3]

Brain2image: Converting brain signals into images,

I. Kavasidis, S. Palazzo, C. Spampinato, D. Giordano, and M. Shah, “Brain2image: Converting brain signals into images,” inProceedings of the 25th ACM International Conference on Multimedia, MM ’17, (New York, NY , USA), p. 1809–1817, Association for Computing Machinery, 2017

work page 2017
[4]

High-resolution image synthesis with latent diffusion models,

R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High-resolution image synthesis with latent diffusion models,” in2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 10674–10685, 2022

work page 2022
[5]

High-resolution image reconstruction with latent diffusion models from human brain activity,

Y . Takagi and S. Nishimoto, “High-resolution image reconstruction with latent diffusion models from human brain activity,” in2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 14453–14463, 2023

work page 2023
[6]

Minddiffuser: Controlled image reconstruction from human brain activity with semantic and structural diffusion,

Y . Lu, C. Du, Q. Zhou, D. Wang, and H. He, “Minddiffuser: Controlled image reconstruction from human brain activity with semantic and structural diffusion,” inProceedings of the 31st ACM International Con- ference on Multimedia, MM ’23, (New York, NY , USA), p. 5899–5908, Association for Computing Machinery, 2023

work page 2023
[7]

Contrast, attend and diffuse to decode high-resolution images from brain activities,

J. Sun, M. Li, Z. Chen, Y . Zhang, S. Wang, and M.-F. Moens, “Contrast, attend and diffuse to decode high-resolution images from brain activities,” inAdvances in Neural Information Processing Systems (A. Oh, T. Naumann, A. Globerson, K. Saenko, M. Hardt, and S. Levine, eds.), vol. 36, pp. 12332–12348, Curran Associates, Inc., 2023

work page 2023

Show all 24 references

[8]

Prolific- dreamer: High-fidelity and diverse text-to-3d generation with variational score distillation,

Z. Wang, C. Lu, Y . Wang, F. Bao, C. LI, H. Su, and J. Zhu, “Prolific- dreamer: High-fidelity and diverse text-to-3d generation with variational score distillation,” inAdvances in Neural Information Processing Sys- tems(A. Oh, T. Naumann, A. Globerson, K. Saenko, M. Hardt, and...

2023
[9]

Dreamfusion: Text- to-3d using 2d diffusion,

B. Poole, A. Jain, J. T. Barron, and B. Mildenhall, “Dreamfusion: Text- to-3d using 2d diffusion,”arXiv preprint arXiv:2209.14988, 2022

2022 arXiv
[10]

Mind-3d: Reconstruct high-quality 3d objects in human brain,

J. Gao, Y . Fu, Y . Wang, X. Qian, J. Feng, and Y . Fu, “Mind-3d: Reconstruct high-quality 3d objects in human brain,” inEuropean Conference on Computer Vision, pp. 312–329, Springer, 2024

2024
[11]

Nerf: representing scenes as neural radiance fields for view synthesis,

B. Mildenhall, P. P. Srinivasan, M. Tancik, J. T. Barron, R. Ramamoorthi, and R. Ng, “Nerf: representing scenes as neural radiance fields for view synthesis,”Commun. ACM, vol. 65, p. 99–106, Dec. 2021

2021
[13]

LoRA: Low-rank adaptation of large language models,

E. J. Hu, Y . Shen, P. Wallis, Z. Allen-Zhu, Y . Li, S. Wang, L. Wang, and W. Chen, “LoRA: Low-rank adaptation of large language models,” inInternational Conference on Learning Representations, 2022

2022
[14]

Decoupled weight decay regularization,

I. Loshchilov and F. Hutter, “Decoupled weight decay regularization,” arXiv preprint arXiv:1711.05101, 2019

2019 arXiv
[15]

A gan model encoded by capseegnet for visual eeg encoding and image reproduction,

X. Deng, Z. Wang, K. Liu, and X. Xiang, “A gan model encoded by capseegnet for visual eeg encoding and image reproduction,”Journal of Neuroscience Methods, vol. 384, p. 109747, 2023

2023
[16]

Visual decoding and reconstruction via eeg embeddings with guided diffusion,

D. Li, C. Wei, S. Li, J. Zou, and Q. Liu, “Visual decoding and reconstruction via eeg embeddings with guided diffusion,” inAdvances in Neural Information Processing Systems(A. Globerson, L. Mackey, D. Belgrave, A. Fan, U. Paquet, J. Tomczak, and C. Zhang, eds.), vol. 37, pp. 1...

2024
[17]

Lstm: A search space odyssey,

K. Greff, R. K. Srivastava, J. Koutn ´ık, B. R. Steunebrink, and J. Schmid- huber, “Lstm: A search space odyssey,”IEEE Transactions on Neural Networks and Learning Systems, vol. 28, no. 10, pp. 2222–2232, 2017

2017
[18]

Generative adversarial networks,

I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y . Bengio, “Generative adversarial networks,” Commun. ACM, vol. 63, p. 139–144, Oct. 2020

2020
[19]

Masked au- toencoders are scalable vision learners,

K. He, X. Chen, S. Xie, Y . Li, P. Doll ´ar, and R. Girshick, “Masked au- toencoders are scalable vision learners,” in2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 15979–15988, 2022

2022
[20]

Decoding Natural Images from EEG for Object Recognition,

Y . Song, B. Liu, X. Li, N. Shi, Y . Wang, and X. Gao, “Decoding Natural Images from EEG for Object Recognition,” inInternational Conference on Learning Representations, 2024

2024
[21]

SDEdit: Guided image synthesis and editing with stochastic differential equations,

C. Meng, Y . He, Y . Song, J. Song, J. Wu, J.-Y . Zhu, and S. Ermon, “SDEdit: Guided image synthesis and editing with stochastic differential equations,” inInternational Conference on Learning Representations, 2022

2022
[22]

Zero-1-to-3: Zero-shot one image to 3d object,

R. Liu, R. Wu, B. V . Hoorick, P. Tokmakov, S. Zakharov, and C. V on- drick, “Zero-1-to-3: Zero-shot one image to 3d object,”arXiv preprint arXiv:2303.11328, 2023

2023 arXiv
[23]

Neuro-3d: Towards 3d visual decoding from eeg signals,

Z. Guo, J. Wu, Y . Song, J. Bu, W. Mai, Q. Zheng, W. Ouyang, and C. Song, “Neuro-3d: Towards 3d visual decoding from eeg signals,” arXiv preprint arXiv:2411.122482024

arXiv
[24]

EEG-driven 3d object recon- struction with style consistency and diffusion prior,

X. Xiang, W. Zhou, and G. Dai, “EEG-driven 3d object recon- struction with style consistency and diffusion prior,”arXiv preprint arXiv:2410.209812024

arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [12]

MOABB: trustworthy algorithm bench- marking for bcis,

V . Jayaram and A. Barachant, “MOABB: trustworthy algorithm bench- marking for bcis,”Journal of Neural Engineering, vol. 15, p. 066011, Sept. 2018

work page 2018

[2] [1]

The human brain in depth: how we see in 3d,

A. E. Welchman, “The human brain in depth: how we see in 3d,”Annual Review of Vision Science, vol. 2, no. 1, pp. 345–376, 2016

work page 2016

[3] [2]

Dreamdiffusion: Generating high-quality images from brain eeg signals,

Y . Bai, X. Wang, Y . Cao, Y . Ge, C. Yuan, and Y . Shan, “Dreamdiffusion: Generating high-quality images from brain eeg signals,”arXiv preprint arXiv:2306.16934, 2023

arXiv 2023

[4] [3]

Brain2image: Converting brain signals into images,

I. Kavasidis, S. Palazzo, C. Spampinato, D. Giordano, and M. Shah, “Brain2image: Converting brain signals into images,” inProceedings of the 25th ACM International Conference on Multimedia, MM ’17, (New York, NY , USA), p. 1809–1817, Association for Computing Machinery, 2017

work page 2017

[5] [4]

High-resolution image synthesis with latent diffusion models,

R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High-resolution image synthesis with latent diffusion models,” in2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 10674–10685, 2022

work page 2022

[6] [5]

High-resolution image reconstruction with latent diffusion models from human brain activity,

Y . Takagi and S. Nishimoto, “High-resolution image reconstruction with latent diffusion models from human brain activity,” in2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 14453–14463, 2023

work page 2023

[7] [6]

Minddiffuser: Controlled image reconstruction from human brain activity with semantic and structural diffusion,

Y . Lu, C. Du, Q. Zhou, D. Wang, and H. He, “Minddiffuser: Controlled image reconstruction from human brain activity with semantic and structural diffusion,” inProceedings of the 31st ACM International Con- ference on Multimedia, MM ’23, (New York, NY , USA), p. 5899–5908, Association for Computing Machinery, 2023

work page 2023

[8] [7]

Contrast, attend and diffuse to decode high-resolution images from brain activities,

J. Sun, M. Li, Z. Chen, Y . Zhang, S. Wang, and M.-F. Moens, “Contrast, attend and diffuse to decode high-resolution images from brain activities,” inAdvances in Neural Information Processing Systems (A. Oh, T. Naumann, A. Globerson, K. Saenko, M. Hardt, and S. Levine, eds.), vol. 36, pp. 12332–12348, Curran Associates, Inc., 2023

work page 2023