Pith. sign in

REVIEW 3 major objections 5 minor 2 cited by

Revisiting Your Memory: Reconstruction of Affect-Contextualized Memory via EEG-guided Audiovisual Generation

T0 review · 3 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read This paper claims that EEG recorded during memory recall can be decoded into per-person affective trajectories with $F_1 = 0.9$, and that these trajectories—converted into text prompts—guide music-video generation that faithfully reflects…

desk verdict New task and dataset, but the F1=0.9 decoding claim is unverified: leave-one-out may only apply to the KNN, not to CEBRA training. read the letter →

arxiv 2412.05296 v2 pith:IVNRRV6T submitted 2024-11-24 cs.AI cs.HCcs.SDeess.AS

Joonwoo Kwon , Heehwan Wang , Jinwoo Lee , Sooyoung Kim , Shinjae Yoo , Yuewei Lin , Jiook Cha This is my paper · ORCID
classification cs.AIcs.HCcs.SDeess.AS
keywords affectivecomputingEEGaffectdecodingautobiographicalmemorymultimodalgenerationpersonalizedmediadynamicsneuralsignalinterpretationaudiovisualsynthesis
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

RevisitAffectiveMemory is a new task: take EEG recorded while someone recalls an autobiographical memory, decode how their momentary feeling state changes across the recall, and use that decoded affect trajectory to generate a music video that re-creates the memory's emotional contour. The authors built a nine-participant dataset in which each person wrote an essay, sketched the scene, picked an accompanying song, and pressed keys to report positive/neutral/negative feelings during recall while EEG was recorded. Their RYM framework extracts per-person valence dynamics from the EEG with a contrastive-learning encoder (reported $F_1 = 0.9$ in leave-one-out decoding), converts the dynamics into affect-word prompts, and lets text-to-music and latent-diffusion image generation produce synchronized audiovisual output. In a user study, videos generated from decoded trajectories matched participants' own session-1 valence reports better than videos from permuted trajectories ($r = 0.265$ vs. $r = 0.095$, $p = .012$), and 56% of participants preferred the decoded-trajectory video. The paper positions this as a step toward personalized, affect-aware media creation driven directly by neural signals.

What carries the argument

The load-bearing component is the affect extractor: a contrastive-learning neural encoder that nonlinearly reduces multi-channel EEG into a low-dimensional latent space in which the auxiliary variable (participant keypress valence) separates the states. The extracted latent valence trajectory is then converted into text prompts through a prompted alignment step using a pre-trained language model plus a fixed emotion-word bank, and these prompts—together with the participant's chosen guiding melody in the case of music—drive a music generation model and a latent diffusion image model to produce the synchronized audiovisual piece.

What would settle it

Retrain the affect encoder with one participant completely excluded (no EEG and no keypress labels from that person) before extracting that participant's trajectory, then recompute the leave-one-out valence $F_1$; if it is much lower than $0.9$, the reported score is an in-sample fit rather than a generalization result. Separately, re-run the session-1 vs session-2 cross-correlation analysis with many random permutations of the decoded trajectory; if the true-condition advantage over permuted trajectories (reported $r = 0.265$ vs $0.095$) does not reliably exceed chance, the alignment between decoded affect and generated content is not established.

Watch

Extended reading notes

Core claim

The central claim is that the temporal affect state during autobiographical recall is decodable from single-trial EEG at the individual level and that this decoded trajectory carries enough information to steer audiovisual generation so the output is affectively aligned with the original memory. The discovery is an application-level demonstration: a contrastive neural encoder trained on multiple sessions of EEG with keypress valence labels produces latent representations whose leave-one-out valence classification reaches $F_1 = 0.9$, and generated music videos conditioned on these decoded trajectories show significantly higher cross-correlation with participants' own affect ratings than videos conditioned on randomly permuted trajectories. The paper treats this as evidence that affective dynamics, rather than only static emotional labels, can be extracted from neural signals and translated into multimodal content.

Load-bearing premise

The load-bearing premise is that the leave-one-out evaluation isolates generalization: the affect decoder must not have seen the held-out participant's EEG or keypress labels when producing that participant's embedding.

Editorial extensions

If this is right

  • Decoding affect dynamics from EEG during natural recall works with a small single-session sample (nine participants), suggesting practical applications where per-user calibration on large datasets is not required.
  • Because generation is driven by a decoded valence trajectory rather than a single emotional label, the framework can render a memory whose feeling state changes over time as a music video with corresponding musical and visual shifts.
  • The session-1-to-session-2 comparison method—measuring cross-correlation between keypress rating trajectories—offers a general template for validating affect-guided generation without requiring ground-truth media, and the reported preference (56%) supports perceived fidelity.
  • The framework's modular design lets the affect-text alignment and generation stages be swapped with any text-conditioned generative model, so improvements in music and video synthesis can inherit into the reconstruction pipeline.

Reading between the lines

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

  • The affect-text alignment bottleneck suggests that the same decoded EEG trajectory could be expressed in other modalities—for example, haptic or olfactory displays—if their prompt interfaces existed, a direction the paper does not explore.
  • A direct extension would be to replace the keypress self-reports with stimulus-derived or video-content labels, which would let the decoder be trained on external affective annotations; this would sacrifice personal idiosyncrasy but would allow application to memories no one rates in real time.
  • The framework does not test whether the decoded trajectories are specific to autobiographical recall or would also be produced by passively watching emotional material; comparing the two settings would clarify what the decoder actually encodes.
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.

Referee Report

3 major / 5 minor

Summary. The paper introduces RevisitAffectiveMemory, a new task and dataset for affect-contextualized reconstruction of autobiographical memories, and proposes RYM, a three-stage pipeline that decodes affect trajectories from EEG with CEBRA, aligns them with text prompts via an LLM, and generates music and video with MusicGen and Stable Diffusion. The dataset contains EEG, text descriptions, sketches, and music from nine participants who recalled mixed-affect memories while providing real-time keypress valence labels. The authors report F1=0.9 for leave-one-out valence decoding and a user study in which videos generated from decoded trajectories correlate more strongly with participants' session-1 valence reports than videos generated from permuted trajectories (r=0.265 vs 0.095, p=.012).

Significance. If the decoding and generation results hold, the paper would make a useful contribution to affective computing by introducing a multimodal dataset and an end-to-end demonstration of EEG-guided personalized media generation. The real-time keypress protocol for capturing affect dynamics during autobiographical recall is a thoughtful design choice, and making the dataset and code available is valuable. However, the central claims currently rest on an evaluation protocol that is not demonstrably out-of-sample. The F1=0.9 result and the user-study correlation can both be explained by label leakage through the CEBRA embeddings. These concerns are fixable with a properly nested cross-validation, but until then the paper's main quantitative claims are unverified.

major comments (3)
  1. [Section 4.2 and Section 5.1] The leave-one-out decoding evaluation as described is not out-of-sample for the CEBRA embeddings. Section 4.2 states that CEBRA was trained with '10 participants' EEG and key-pressed valence sequence' and then describes leave-one-out only for the KNN classifier, whose training uses 9 participants' identified embeddings and evaluation uses the remaining participant's embeddings. Because CEBRA's contrastive objective used the held-out participant's keypress labels to shape that participant's embedding, the KNN evaluation can reflect label leakage rather than generalization of neural decoding. Consequently, the F1=0.9 reported in Section 5.1 does not substantiate the abstract's claim of decoding individual affect trajectories. The authors should retrain CEBRA in each fold with the held-out participant excluded from all training data, report per-fold F1 with mean and standard deviation, and provide a per-class breakdown. They should also resolve the inconsistency between the 9 participants reported in Section 3 and the 10 participants named in Section 4.2.
  2. [Section 5.4] The user-study result is contaminated by the same protocol issue. The 'true' videos were generated from CEBRA-decoded affect sequences, but those embeddings were produced by a model trained on the same participants' keypress labels, so the higher cross-correlation with session-1 keypresses (r_true=0.265 vs r_perm=0.095, p=.012) may reflect memorized label sequences rather than EEG-based decoding. The authors should regenerate the 'true' videos using per-fold retrained, out-of-sample embeddings, or otherwise demonstrate that the advantage persists when the CEBRA encoder is trained without the test participant. In addition, the definition of 'best cross-correlation coefficients' (e.g., maximum over which time lags?) should be stated, and the appropriate test is a paired Wilcoxon signed-rank test rather than a rank-sum test, since each participant contributes one pair of coefficients from the true and permuted conditions.
  3. [Section 4.3] The mapping from decoded affect to text prompt is underspecified and the wording undercuts the contribution. Section 4.3 says 'we randomly selected words from our word bank and used them constructing prompts.' If word selection is random over the full word bank, then the affect extractor's valence labels do not determine the generated content, in which case 'affect-contextualized' is not supported. If words are sampled uniformly within the decoded positive or negative valence category (with neutral prompts unaltered), this should be stated explicitly, together with how the valence trajectory maps to prompt durations and to the crossfaded music segments described in Section 4.4.
minor comments (5)
  1. [Section 3.1] Please specify whether the '1' and '3' keypresses are held continuously or are discrete events, and how the resulting sparse label sequence is converted to timepoint-level labels for CEBRA (e.g., forward-fill until the next press).
  2. [Section 4.4] The sentence 'The neighborhood parameters of KNN classifiers were fixed at the square of the number of the input time points' is ambiguous; presumably k = sqrt(number of time points), and the resulting value should be reported.
  3. [Table 1 and Section 5.3] Provide definitions and units for the CLIP and CLAP distances, and clarify the 'Affect Difference' rows; no statistical comparison is reported, so the claim of semantic coherence is descriptive only.
  4. [Figure 5] Report confidence intervals or statistical tests for the eta-squared effect sizes; with n=9, a descriptive 'larger' comparison is not informative.
  5. [Section 5.4] The preference result should be reported with exact counts (5 out of 9 participants), and the analysis should state how 'Both' and 'Neither' responses are handled; 56% is not a strong majority at n=9.

Circularity Check

2 steps flagged · score 6.0 of 10

CEBRA is trained on all participants' EEG and keypress labels, so the F1=0.9 decoding and the r=0.265 reconstruction advantage are in-sample fits rather than predictions.

  1. fitted input called prediction [Section 4.2 ('Extracting Human Affects') and Section 5.1 ('Affect Decoding')]
    "Our affect extractor utilizes keypressed valence sequences (neutral, positive, negative) as auxiliary variables to identify individual-level latent clusters for each affective state. We performed multi-session training with 10 participants' EEG and key-pressed valence sequence... We trained a k-nearest neighbor (KNN) classifier with 9 participants' identified embeddings to predict valence label at each timepoint and evaluated its predictive performance with the other's embeddings. ..."

    The supervised auxiliary variable used to train CEBRA is the keypressed valence label, and the same labels are the prediction target. The leave-one-out procedure is applied only to the KNN classifier, not to CEBRA. Because CEBRA was trained via 'multi-session training with 10 participants' EEG and key-pressed valence sequence,' the held-out participant's embedding was produced by a model that already saw that participant's EEG and keypress labels. The F1=0.9 therefore measures how well the embedding memorizes its own training labels, not how well neural signals generalize to a new participant. Retraining CEBRA per fold with the held-out participant excluded would be required for the claimed decoding result.

  2. fitted input called prediction [Section 3.2 ('Experiment Session 2') and Section 5.4 ('User Study')]
    "To rigorously evaluate whether the contents generated with RYM properly reflect subjective affect dynamics, we compared 'real' (based on their CEBRA-decoded affective sequence) against 'fake' (based on permuted affective sequences) music videos. ... the true condition produced significantly higher coefficients than the perm condition (r_true = 0.265, r_perm = 0.095, p = .012)."

    The 'real' videos are generated from the CEBRA-decoded affective sequence, which is shaped by the session-1 keypress labels used in CEBRA training. Comparing session-2 keypresses against session-1 keypresses then tests whether the generated video conveys the order of the original keypresses, not whether EEG independently decoded affect. The 'fake' condition shuffles only the temporal order of the same label-derived trajectory, leaving the label content intact, so it cannot control for the label leakage in the CEBRA embeddings. The r=0.265 advantage is thus partially forced by construction.

full rationale

The central derivation chain reduces to the keypress labels at two points. First, Section 4.2 states that CEBRA's auxiliary variables are the keypressed valence sequences and that 'multi-session training' used 10 participants' EEG and keypress data; the leave-one-out step is described only for the KNN classifier. Since CEBRA is a supervised contrastive encoder, the held-out participant's embedding is a function of that participant's own labels, so the reported F1=0.9 is an in-sample fit. Second, Section 5.4's real-versus-fake comparison inherits this leakage: the 'real' videos are generated from the leaked embeddings, so the higher cross-correlation with session-1 keypresses partially reflects the pipeline's access to those labels, while the permuted baseline only destroys temporal order. The 9-versus-10 participant inconsistency in the text further obscures whether a per-fold retraining was performed; as written, no per-fold CEBRA retraining is described. The self-citation of [53] for hyperparameters exists but is not the main circularity because CEBRA itself is externally published. The quantitative CLIP/CLAP and feature analyses are self-contained and do not depend on the leaked decoding, but they do not rescue the paper's central neural-decoding and reconstruction claims.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim depends on self-reported keypress labels as ground truth, on cross-subject alignment of CEBRA embeddings, and on lossless conversion of decoded affect into text prompts. Several hyperparameters (embedding dimensionality, CEBRA settings, KNN k) are chosen by hand or from prior same-lab work and are not ablated. No new physical or theoretical entities are introduced.

free parameters (3)
  • CEBRA latent embedding dimensionality = 7
    Set to 7 based on a prior same-lab VR-EEG study [53]; no ablation or analysis for this dataset. The reported F1=0.9 depends on this choice.
  • CEBRA training hyperparameters = batch_size=2048, hidden_units=95, learning_rate=0.005, iterations=2000, offset-10 model, hybrid=False
    Chosen by hand or taken from previous work; no sensitivity analysis is reported. These settings shape the embeddings used in decoding and generation.
  • KNN neighborhood parameter = k = sqrt(number of input time points)
    Fixed at the square of the number of input time points in Section 4.4; no analysis of how this affects the F1 score.
assumptions (4)
  • domain assumption Keypress valence labels (1=positive, 3=negative, no key=neutral) accurately reflect participants' affective states during memory recall.
    Used as ground truth for CEBRA training and for evaluating generated videos. Discrete self-report may miss mixed or ambivalent affect; introduced in Section 3.1.
  • domain assumption CEBRA embeddings trained on a set of participants align across individuals, so a KNN trained on nine participants' embeddings can classify a held-out participant.
    The leave-one-out protocol in Section 4.2 assumes cross-subject alignment of latent representations; evidence is limited to qualitative plots in Figure 3c.
  • domain assumption Text prompts produced by Claude 3.5 Sonnet from memory essays and emotion words preserve the decoded affective information for downstream generative models.
    The affect-to-text alignment in Section 4.3 is assumed lossless; no direct check that prompts reflect intended affect beyond CLIP/CLAP distances.
  • ad hoc to paper The 7-dimensional embedding setting from a prior same-lab study transfers to autobiographical memory recall.
    Section 4.4 imports this design choice from Yi et al. 2024 [53] without an ablation; the decoding performance is conditional on it.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Revisiting Your Memory: Reconstruction of Affect-Contextualized Memory via EEG-guided Audiovisual Generation." pith.science (2026). https://pith.science/paper/IVNRRV6T

@misc{pith2026241205296,
  author       = {Pith},
  title        = {Pith review of: Revisiting Your Memory: Reconstruction of Affect-Contextualized Memory via EEG-guided Audiovisual Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IVNRRV6T}},
  note         = {Machine review of arXiv:2412.05296}
}
read the original abstract

In this paper, we introduce RevisitAffectiveMemory, a novel task designed to reconstruct autobiographical memories through audio-visual generation guided by affect extracted from electroencephalogram (EEG) signals. To support this pioneering task, we present the EEG-AffectiveMemory dataset, which encompasses textual descriptions, visuals, music, and EEG recordings collected during memory recall from nine participants. Furthermore, we propose RYM (Revisit Your Memory), a three-stage framework for generating synchronized audio-visual contents while maintaining dynamic personal memory affect trajectories. Experimental results demonstrate our method successfully decodes individual affect dynamics trajectories from neural signals during memory recall (F1=0.9). Also, our approach faithfully reconstructs affect-contextualized audio-visual memory across all subjects, both qualitatively and quantitatively, with participants reporting strong affective concordance between their recalled memories and the generated content. Especially, contents generated from subject-reported affect dynamics showed higher correlation with participants' reported affect dynamics trajectories (r=0.265, p<.05) and received stronger user preference (preference=56%) compared to those generated from randomly reordered affect dynamics. Our approaches advance affect decoding research and its practical applications in personalized media creation via neural-based affect comprehension. Codes and the dataset are available at https://github.com/ioahKwon/Revisiting-Your-Memory.

Figures

Figures reproduced from arXiv: 2412.05296 by the authors.

Figure 1
Figure 1. Study overview. (a) Conceptual scheme of RevisitAffectiveMemory task. (b) The EEG-AffectiveMemory dataset [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The RYM architecture comprises three primary components: an affect extractor, an affect-text alignment, and affect [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Our affect decoding results. (a) idiosyncratic temporal dynamics of real-time valence keypress during memory recall [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Qualitative results. (a) the images and (b) the music (mel-spectogram for visualization) are randomly selected. Note [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: The affective correlations of visual and audio attributes. (a) Comparison of true and fake (i.e., random) music videos’ [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: User study results. (a) Cross-correlation coefficients [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. WorldWeaver: Generating Long-Horizon Video Worlds via Rich Perception

    cs.CV 2025-08 unverdicted novelty 5.0 of 10

    WorldWeaver reduces temporal drift in long-horizon video generation by jointly modeling RGB and depth perceptual conditions with segmented noise scheduling.

  2. Foundation Models for Cross-Domain EEG Analysis Application: A Survey

    cs.HC 2025-08 conditional novelty 4.0 of 10

    A survey that organizes EEG foundation-model research into five output-modality categories: native EEG, text, vision, audio, and multimodal fusion, with a claim to be the first such comprehensive taxonomy.

Reference graph

Works this paper leans on

58 extracted references · 38 canonical work pages · cited by 2 Pith papers

  1. [1]

    Andrea Agostinelli, Timo I Denk, Zalán Borsos, Jesse Engel, Mauro Verzetti, An- toine Caillon, Qingqing Huang, Aren Jansen, Adam Roberts, Marco Tagliasacchi, et al. 2023. Musiclm: Generating music from text. arXiv preprint arXiv:2301.11325 (2023)

    arXiv 2023

  2. [2]

    Abeer Al-Nafjan, Manar Hosny, Yousef Al-Ohali, and Areej Al-Wabil. 2017. Re- view and classification of emotion recognition based on EEG brain-computer interface system research: a systematic review. Applied Sciences 7, 12 (2017), 1239

    work page 2017

  3. [3]

    Prashanth Chandran, Gaspard Zoss, Paulo Gotardo, Markus Gross, and Derek Bradley. 2021. Adaptive convolutions for structure-aware style transfer. In Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition . 7972–7981

    work page 2021

  4. [4]

    Guillaume Chanel, Joep JM Kierkels, Mohammad Soleymani, and Thierry Pun

    work page

  5. [5]

    Jin Chen, Tony Ro, and Zhigang Zhu. 2022. Emotion recognition with audio, video, EEG, and EMG: a dataset and baseline approaches. IEEE Access 10 (2022), 13229–13242

    work page 2022

  6. [6]

    Tian Qi Chen and Mark Schmidt. 2016. Fast patch-based style transfer of arbitrary style. arXiv preprint arXiv:1612.04337 (2016)

    arXiv 2016

  7. [7]

    Jiwoo Chung, Sangeek Hyun, and Jae-Pil Heo. 2024. Style injection in diffusion: A training-free approach for adapting large-scale diffusion models for style transfer. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 8795–8805

    2024

  8. [8]

    Ondřej Cífka, Umut Şimşekli, and Gaël Richard. 2019. Supervised symbolic music style translation using synthetic data. arXiv preprint arXiv:1907.02265 (2019)

    arXiv 2019

Show all 58 references
  1. [9]

    Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi, and Alexandre Défossez. 2024. Simple and controllable music generation. Advances in Neural Information Processing Systems 36 (2024)

    2024

  2. [10]

    Shuqi Dai, Zheng Zhang, and Gus G Xia. 2018. Music style transfer: A position paper. arXiv preprint arXiv:1803.06841 (2018)

    2018 arXiv

  3. [11]

    Muhammad Najam Dar, Muhammad Usman Akram, Ahmad Rauf Subhani, Sajid Gul Khawaja, Constantino Carlos Reyes-Aldasoro, and Sarah Gul. 2024. Insights from EEG analysis of evoked memory recalls using deep learning for emotion charting. Scientific Reports 14, 1 (2024), 17080

    2024

  4. [12]

    Jesse Engel, Cinjon Resnick, Adam Roberts, Sander Dieleman, Mohammad Norouzi, Douglas Eck, and Karen Simonyan. 2017. Neural audio synthesis of musical notes with wavenet autoencoders. InInternational Conference on Machine Learning. PMLR, 1068–1077

    2017

  5. [13]

    Martin Nicolas Everaert, Marco Bocchio, Sami Arpa, Sabine Süsstrunk, and Radhakrishna Achanta. 2023. Diffusion in style. In Proceedings of the IEEE/CVF International Conference on Computer Vision . 2251–2261

    2023

  6. [14]

    Leon A Gatys, Alexander S Ecker, and Matthias Bethge. 2016. Image style transfer using convolutional neural networks. In Proceedings of the IEEE conference on computer vision and pattern recognition . 2414–2423

    2016

  7. [15]

    Eric Grinstein, Ngoc QK Duong, Alexey Ozerov, and Patrick Pérez. 2018. Audio style transfer. In 2018 IEEE international conference on acoustics, speech and signal processing (ICASSP). IEEE, 586–590

    2018

  8. [16]

    Jing Huang, Shizhe Zhou, Xianyi Zhu, Yiwen Li, and Chengfeng Zhou. 2018. Automatic image style transfer using emotion-palette. In Tenth International Conference on Digital Image Processing (ICDIP 2018) , Vol. 10806. SPIE, 1197–1206

    2018

  9. [17]

    Sicong Huang, Qiyang Li, Cem Anil, Xuchan Bao, Sageev Oore, and Roger B Grosse. 2018. Timbretron: A wavenet (cyclegan (cqt (audio))) pipeline for musical timbre transfer. arXiv preprint arXiv:1811.09620 (2018)

    2018 arXiv

  10. [18]

    Xun Huang and Serge Belongie. 2017. Arbitrary style transfer in real-time with adaptive instance normalization. In Proceedings of the IEEE international conference on computer vision . 1501–1510

    2017

  11. [19]

    Daniela Iacoviello, Andrea Petracca, Matteo Spezialetti, and Giuseppe Placidi

  12. [20]

    Lutz Jäncke. 2008. Music, memory and emotion. Journal of biology 7 (2008), 1–5

    2008

  13. [21]

    Justin Johnson, Alexandre Alahi, and Li Fei-Fei. 2016. Perceptual losses for real- time style transfer and super-resolution. In Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceed- ings, Part II 14 . Springer, 694–711

    2016

  14. [22]

    Eshin Jolly, Max Farrens, Nathan Greenstein, Hedwig Eisenbarth, Marianne C Reddan, Eric Andrews, Tor D Wager, and Luke J Chang. 2022. Recovering individual emotional states from sparse ratings using collaborative filtering. Affective Science 3, 4 (2022), 799–817

    2022

  15. [23]

    Stamos Katsigiannis and Naeem Ramzan. 2017. DREAMER: A database for emotion recognition through EEG and ECG signals from wireless low-cost off- the-shelf devices. IEEE journal of biomedical and health informatics 22, 1 (2017), 98–107

    2017

  16. [24]

    Sooyoung Kim, Joonwoo Kwon, Heehwan Wang, Shinjae Yoo, Yuewei Lin, and Jiook Cha. 2024. A Training-Free Approach for Music Style Transfer with Latent Diffusion Models. arXiv preprint arXiv:2411.15913 (2024)

    2024 arXiv

  17. [25]

    Sander Koelstra, Christian Muhl, Mohammad Soleymani, Jong-Seok Lee, Ashkan Yazdani, Touradj Ebrahimi, Thierry Pun, Anton Nijholt, and Ioannis Patras. 2011. Deap: A database for emotion analysis; using physiological signals. IEEE transac- tions on affective computing 3, 1 (2011), 18–31

    2011

  18. [26]

    Dmytro Kotovenko, Artsiom Sanakoyeu, Sabine Lang, and Bjorn Ommer. 2019. Content and style disentanglement for artistic style transfer. In Proceedings of the IEEE/CVF international conference on computer vision . 4422–4431

    2019

  19. [27]

    Peter Kuppens and Philippe Verduyn. 2017. Emotion dynamics. Current Opinion in Psychology 17 (2017), 22–26

    2017

  20. [28]

    Gihyun Kwon and Jong Chul Ye. 2022. Clipstyler: Image style transfer with a single text condition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition . 18062–18071

    2022

  21. [29]

    Joonwoo Kwon, Sooyoung Kim, Yuewei Lin, Shinjae Yoo, and Jiook Cha. 2024. Aesfa: an aesthetic feature-aware arbitrary neural style transfer. In Proceedings of the AAAI conference on artificial intelligence , Vol. 38. 13310–13319

    2024

  22. [30]

    Pilhyeon Lee, Sunhee Hwang, Seogkyu Jeon, and Hyeran Byun. 2021. Subject adaptive eeg-based visual recognition. In Asian Conference on Pattern Recognition. Springer, 322–334

    2021

  23. [31]

    Kai Liu, Wei Li, Lai Chen, Shengqiong Wu, Yanhao Zheng, Jiayi Ji, Fan Zhou, Rongxin Jiang, Jiebo Luo, Hao Fei, et al. 2025. Javisdit: Joint audio-video diffusion transformer with hierarchical spatio-temporal prior synchronization. arXiv preprint arXiv:2503.23377 (2025)

    2025

  24. [32]

    Mason McClay, Matthew E Sachs, and David Clewett. 2023. Dynamic emotional states shape the episodic structure of memory. Nature Communications 14, 1 (2023), 6533

    2023

  25. [33]

    Timon Merk, Richard Köhler, Victoria Peterson, Laura Lyra, Jonathan Vanhoecke, Meera Chikermane, Thomas Binns, Ningfei Li, Ashley Walton, Alan Bush, et al

  26. [34]

    Caitlin Mills and Sidney D’Mello. 2014. On the validity of the autobiographical emotional memory task for emotion induction. PloS one 9, 4 (2014), e95837

    2014

  27. [35]

    Juan Abdon Miranda-Correa, Mojtaba Khomami Abadi, Nicu Sebe, and Ioannis Patras. 2018. Amigos: A dataset for affect, personality and mood research on individuals and groups. IEEE transactions on affective computing 12, 2 (2018), 479–493

    2018

  28. [36]

    In-Jo Park and Kyunghwan Min. 2005. Making a list of Korean emotion terms and exploring dimensions underlying them. Korean Journal of Social and Personality Psychology 19, 1 (2005), 109–129. CogMAEC ’25, October 27–31, 2025, Dublin, Ireland. Kwon et al

    2005

  29. [37]

    Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, et al. 2019. Pytorch: An imperative style, high-performance deep learning library. Advances in neural information processing system...

    2019

  30. [38]

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Björn Ommer. 2022. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 10684–10695

    2022

  31. [39]

    Ludan Ruan, Yiyang Ma, Huan Yang, Huiguo He, Bei Liu, Jianlong Fu, Nicholas Jing Yuan, Qin Jin, and Baining Guo. 2023. Mm-diffusion: Learning multi-modal diffusion models for joint audio and video generation. In Proceed- ings of the IEEE/CVF Conference on Computer Vision and P...

    2023

  32. [40]

    Heini Saarimäki. 2021. Naturalistic stimuli in affective neuroimaging: A review. Frontiers in human neuroscience 15 (2021), 675068

    2021

  33. [41]

    Steffen Schneider, Jin Hwa Lee, and Mackenzie Weygandt Mathis. 2023. Learnable latent embeddings for joint behavioural and neural analysis. Nature 617, 7960 (2023), 360–368

    2023

  34. [42]

    Hirokatsu Shimizu and Ramesh Srinivasan. 2022. Improving classification and reconstruction of imagined images from EEG signals. Plos one 17, 9 (2022), e0274847

    2022

  35. [43]

    Ewa Siedlecka and Thomas F Denson. 2019. Experimental methods for inducing basic emotions: A qualitative review. Emotion Review 11, 1 (2019), 87–97

    2019

  36. [44]

    Miles Sigel, Michael Zhou, and Jiebo Luo. 2021. Music Sentiment Transfer. arXiv preprint arXiv:2110.05765 (2021)

    2021 arXiv

  37. [45]

    Mingzhen Sun, Weining Wang, Yanyuan Qiao, Jiahui Sun, Zihan Qin, Longteng Guo, Xinxin Zhu, and Jing Liu. 2024. Mm-ldm: Multi-modal latent diffusion model for sounding video generation. In Proceedings of the 32nd ACM International Conference on Multimedia. 10853–10861

    2024

  38. [46]

    Hao Hao Tan and Dorien Herremans. 2020. Music fadernets: Controllable music generation based on high-level features via low-level feature modelling. arXiv preprint arXiv:2007.15474 (2020)

    2020 arXiv

  39. [47]

    Dmitry Ulyanov, Vadim Lebedev, Andrea Vedaldi, and Victor Lempitsky. 2016. Texture networks: Feed-forward synthesis of textures and stylized images. arXiv preprint arXiv:1603.03417 (2016)

    2016 arXiv

  40. [48]

    Anthony G Vaccaro, Helen Wu, Rishab Iyer, Shruti Shakthivel, Nina C Christie, Antonio Damasio, and Jonas Kaplan. 2024. Neural patterns associated with mixed valence feelings differ in consistency and predictability throughout the brain. Cerebral Cortex 34, 4 (2024), bhae122

    2024

  41. [49]

    R Vempati and L Sharma. 2023. A systematic review on automated human emotion recognition using electroencephalogram signals and artificial intelligence. Results Eng. 18, 101027 (2023)

    2023

  42. [50]

    Zhizhong Wang, Lei Zhao, and Wei Xing. 2023. Stylediffusion: Controllable disentangled style transfer via diffusion models. In Proceedings of the IEEE/CVF International Conference on Computer Vision . 7677–7689

    2023

  43. [51]

    Zhizhong Wang, Lei Zhao, Zhiwen Zuo, Ailin Li, Haibo Chen, Wei Xing, and Dongming Lu. 2023. MicroAST: Towards Super-Fast Ultra-Resolution Arbitrary Style Transfer. In Proceedings of the AAAI Conference on Artificial Intelligence

    2023

  44. [52]

    Shuchen Weng, Peixuan Zhang, Zheng Chang, Xinlong Wang, Si Li, and Boxin Shi. 2023. Affective image filter: Reflecting emotions from text to images. In Proceedings of the IEEE/CVF International Conference on Computer Vision . 10810– 10819

    2023

  45. [53]

    Jinwoo Yi, Danny Dongyeop Han, Seung-Yeop Oh, and Jiook Cha. 2024. Awe is characterized as an ambivalent experience in the human behavior and cortex: integrated virtual reality-electroencephalogram study. bioRxiv (2024), 2024–08

    2024

  46. [54]

    Yuxin Zhang, Nisha Huang, Fan Tang, Haibin Huang, Chongyang Ma, Weiming Dong, and Changsheng Xu. 2023. Inversion-based style transfer with diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 10146–10156

    2023

  47. [55]

    Wei-Long Zheng and Bao-Liang Lu. 2015. Investigating critical frequency bands and channels for EEG-based emotion recognition with deep neural networks. IEEE Transactions on autonomous mental development 7, 3 (2015), 162–175. Received 30 June 2025; accepted 5 August 2025

    2015

  48. [2009]

    International Journal of Human-Computer Studies 67, 8 (2009), 607–627

    Short-term emotion assessment in a recall paradigm. International Journal of Human-Computer Studies 67, 8 (2009), 607–627

    2009

  49. [2015]

    Computer methods and programs in biomedicine 122, 3 (2015), 293–303

    A real-time classification algorithm for EEG-based BCI driven by self- induced emotions. Computer methods and programs in biomedicine 122, 3 (2015), 293–303

    2015

  50. [2023]

    Research Square (2023)

    Invasive neurophysiology and whole brain connectomics for neural decod- ing in patients with brain implants. Research Square (2023)

    2023

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.