Pith. sign in

REVIEW 5 major objections 5 minor 44 references

Sound-Based Recognition of Touch Gestures and Emotions for Enhanced Human-Robot Interaction

T0 review · 5 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read This paper shows that the sound of touching a robot carries enough signal to recognize six touch gestures and to classify emotions by arousal and valence, using a 0.24-million-parameter audio model that can run on Pepper's onboard computer.

desk verdict Novel and motivated idea for audio-only touch gesture and emotion recognition on a robot, but the reported accuracies likely leak participant identity because the paper never states that the train/test split is participant-independent. read the letter →

arxiv 2501.00038 v1 pith:67UXD7FD submitted 2024-12-24 cs.HC cs.ROcs.SDeess.AS

Yuanbo Hou , Qiaoqiao Ren , Wenwu Wang , Dick Botteldooren This is my paper · ORCID
classification cs.HCcs.ROcs.SDeess.AS
keywords touchgesturerecognitionemotionarousal-valencemodelhuman-robotinteractionaudio-onlyclassificationmulti-temporal-resolutionCNNPepperrobotMTRCNN
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

The paper's goal is to give humanoid robots a privacy-friendly way to feel touch: instead of installing full-body tactile skin or relying on cameras that raise data-protection concerns, it asks whether the sounds people make when touching the robot can reveal both the gesture and the emotion behind it. The authors propose MTRCNN, a multi-temporal-resolution convolutional network with 0.24 million parameters, 0.94 MB size, and 0.708 G FLOPs, and evaluate it on recordings of 28 participants touching the Pepper robot's forearm. They report that the model distinguishes six touch gestures and places emotions in an arousal-valence space, with accuracy that holds when the input clip varies from about 2 seconds up to an optimum of 6–7 seconds. It matters because this is the first attempt to decode both gesture and emotion from the non-speech sounds of touch during human-robot interaction, and the model's size and latency fit the constraints of Pepper's onboard brain.

What carries the argument

The carrying object is MTRCNN, a multi-temporal-resolution convolutional neural network. It runs three parallel convolutional branches with kernel sizes (3,3), (5,5), and (7,7), applies hybrid dilated convolutions with dilation rates (1,1), (2,1), and (3,1) along the time axis to enlarge the receptive field, pools so that the minimum input is 1.10 seconds, and fuses three 64-dimensional embeddings into a 192-dimensional representation. Four softmax heads consume that representation for arousal, valence, joint arousal-valence, and gesture classification; the key property is that the small kernel set and dilated time-axis filters let a 0.24M-parameter model capture long-range rhythm and contact patterns in variable-length clips.

What would settle it

Rerun MTRCNN under a strict leave-one-participant-out protocol: train on clips from 27 participants and test on the held-out 28th participant. If six-way gesture accuracy falls from the reported 82-90% toward the 17% chance level, or five-way arousal-valence accuracy falls toward the 20% chance level, then the model is learning participant-specific sound profiles rather than the named gestures and emotion states.

Watch

Extended reading notes

Core claim

The central claim, stated on the paper's own terms, is that touch sounds are informative enough to support joint gesture and emotion recognition on a real robot platform. On 10-second clips the model reaches 82.14% accuracy across the six gestures and 53.93% across the five joint arousal-valence categories; on the validation set the best performance is 90.24% for gestures with 6-second clips and 58.54% for arousal-valence with 7-second clips, with paired t-tests reported as confirming that those lengths beat neighboring clip lengths. The same 0.24M-parameter network achieves higher gesture accuracy than PANNs pretrained on AudioSet, and is not significantly different from it on arousal-valence, while using far fewer parameters, FLOPs, and memory.

Load-bearing premise

The reported accuracies assume that clips from the same person never appear in both the training and the test sets, but the paper does not say whether the data were split that way.

Editorial extensions

If this is right

  • A robot with a microphone but no tactile skin can classify six touch gestures in real time, since inference on a clip takes about 7 milliseconds.
  • Because the model needs only 0.708 G FLOPs, it fits the computing ceiling of Pepper's onboard processor and can run on-device without cloud support.
  • Large-scale audio pretraining is not required for this task: the PANNs baseline pretrained on AudioSet is not significantly better than MTRCNN on arousal-valence joint classification, and MTRCNN is better on gesture classification.
  • The effective interaction window is about 6 seconds for gestures and 7 seconds for emotions, and clips down to about 2 seconds still work, so future deployments can segment interaction audio accordingly.

Reading between the lines

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

  • The paper does not state whether the train/validation/test split keeps each participant's clips in a single partition; with 28 participants and three clips each, a clip-level random split would almost surely place the same person in both training and test, so a strict leave-one-participant-out evaluation is needed to confirm that the model generalizes to new people.
  • If the split-leakage risk is real, the absolute accuracies in Tables I and III are likely optimistic, although the relative ordering of clip lengths and model comparisons would still be informative.
  • The same architecture and touch-sound idea could plausibly extend to other non-verbal contact sounds in human-robot interaction, such as footsteps, handshakes, or object taps, where on-robot microphones already record audio; the paper does not test this.
  • A natural next test is to feed a continuous interaction stream and let the model segment its own windows, since the paper evaluates only pre-segmented clips and the optimal 6–7 second length may depend on how the clips were cut.
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

5 major / 5 minor

Summary. The paper proposes a lightweight audio-only model (MTRCNN) that recognizes touch gestures (hold, pat, poke, tickle, tap, rub) and emotion arousal/valence states from sounds produced during human-robot interaction with the Pepper robot. The model uses multi-branch dilated convolutions and a fusion layer, with 0.24M parameters and 0.7G FLOPs. Experiments on a dataset of 28 participants report test accuracies around 82% for gestures and 54% for joint arousal-valence classification, with an analysis claiming optimal input lengths of 6–7 seconds. The paper compares MTRCNN against several larger baselines, including pretrained PANNs, and concludes that the model is suitable for real-time deployment on Pepper.

Significance. If the reported results hold, this is a valuable contribution to privacy-preserving HRI, offering a low-cost, non-intrusive alternative to vision- or tactile-sensor-based emotion and gesture recognition. The paper provides a dataset, code, and models, which supports reproducibility and follow-up research. The lightweight architecture and low latency are concrete strengths for embedded deployment. However, the evaluation has several gaps that currently undermine the central quantitative claims, especially the unclear split protocol and missing input-length specification in comparisons. The practical significance is high if these issues are resolved.

major comments (5)
  1. [Section IV-A] The train/validation/test split protocol is not specified at the participant level. With 28 participants each contributing three clips per gesture/emotion, a random clip-level split will almost certainly place clips from the same participant in both training and test, allowing the model to exploit participant-specific recording conditions or hand acoustics. This would inflate all reported accuracies in Tables I and III, as well as the validation-based conclusions in Table II. The authors must state explicitly whether all clips from a given participant are confined to a single partition, and if not, re-run the experiments with a participant-independent split.
  2. [Section IV-A] The gesture clip counts are inconsistent. For six gestures with 84 clips each, the total should be 504 clips, but the reported training/validation/test numbers sum to 366+42+84=492, leaving 12 clips unaccounted for. The manuscript should explain how these clips were discarded and confirm that the remaining clips are balanced across gestures and participants. Without this, the reported gesture accuracy may not reflect the full dataset.
  3. [Table III] Table III does not state the input audio length used for each model. Since the paper concludes in RQ2 that 6s and 7s are the optimal lengths for gesture and Aro-Val classification and declares these as defaults, Table III must specify the input length for each method and use a consistent length across models. The current ambiguity is material: MTRCNN's gesture accuracy in Table III (84.17%) differs from the 10s result in Table I (82.14%), so it is unclear whether the table uses the optimal lengths or something else.
  4. [Section IV-B, RQ2] The optimal-length analysis in Table II is performed only on the validation set, and the selected lengths (6s for gestures, 7s for Aro-Val) are never evaluated on the held-out test set. The conclusion that these lengths are optimal is therefore not supported by test-set evidence. The authors should report test-set accuracies for the selected lengths, as well as for nearby lengths (e.g., 5s, 6s, 7s, 8s), to confirm the choice on unseen data.
  5. [Section IV-B, RQ2] The paired t-tests comparing accuracies at different input lengths may be statistically invalid. The paper states that models were trained '10 times without a fixed seed,' meaning the runs for different input lengths are independent, not paired. A paired t-test requires paired observations (e.g., identical initialization across conditions). If the runs are independent, an unpaired test or a repeated-measures design with fixed seeds should be used. This affects the significance claims for the optimal length selection.
minor comments (5)
  1. [Section IV-A] The sentence 'participants first expressed gestures 10 emotions independently' is missing a conjunction; it should read 'gestures and 10 emotions.'
  2. [Section IV-B, RQ3] The confusion matrices in Fig. 5 are reported only on the test set, but it is not stated which input length was used for the model when generating them. For consistency with RQ2, the figure should specify whether the default 6s/7s lengths were used.
  3. [Section III-B] The architecture description omits the pooling sizes and the exact placement of pooling operations, despite the minimum input length of 1.10s being derived from pooling. The GitHub link is useful, but the paper should include these details for self-contained reproducibility.
  4. [References] Several references are incomplete or inconsistent, e.g., [12] lacks a volume/issue, and [41] uses only a journal name without volume. Please harmonize the reference style throughout.
  5. [Abstract] The abstract claims the model 'achieves similar results as well-known pretrained audio neural networks (PANNs)' but the statistical test for this claim is only reported for Aro-Val, not for gesture. Clarify the comparison or add a significance test for gesture accuracy.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation found: the model is evaluated on a genuine held-out test set, and the architecture's receptive-field and length limits follow from standard convolution formulas.

full rationale

The paper's central claim is that a lightweight audio model can recognize touch gestures and emotion dimensions from touch sounds recorded during human-robot interaction. The recognition pipeline is a standard supervised learning setup: audio clips are converted to Mel features, a convolutional network with multiple temporal resolutions is designed from receptive-field arithmetic (Eqs. 1 and 2), trained with cross-entropy loss, and evaluated on a test set whose reported accuracies in Tables I and III are genuine model outputs rather than fitted parameters renamed as predictions. The input-length analysis in Table II is explicitly performed on the validation set to select hyperparameters, which is standard practice and does not constitute circularity; the conclusion that 6-7 s is optimal is an empirical observation, not a quantity built into the model definition. The receptive-field calculation that yields the 1.10 s minimum input length is derived from the convolution and dilated-convolution formulas and therefore is not circular. Citations to prior work by the same authors (e.g., [28], [29], [32], [33]) are used for background, for the dataset, and for standard technical components such as cross entropy; none of these is invoked to prove the central claim in a way that reduces to self-citation. The dataset itself is an external empirical resource, and the reported test-set results provide independent evidence for the central claim. Concerns about whether the train/validation/test split is participant-independent would bear on the validity of the reported accuracies, but that is a data-protocol correctness issue, not a circularity of the derivation chain. Overall, no prediction or claimed result is equivalent by construction to its inputs.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The central claim rests on standard machine-learning and affect-modeling assumptions rather than on new physical laws. The key load-bearing choices are the circumplex emotion model, the assumption that touch sounds carry emotion cues, and the unstated split protocol. No invented entities are introduced.

free parameters (3)
  • Default input audio lengths = 6 s for gestures, 7 s for Aro-Val
    Selected as the best lengths on the validation set (Table II) and used for the reported test results; this is a fitted hyperparameter.
  • Architecture hyperparameters = Kernel sizes 3/5/7, dilation rates (1,1),(2,1),(3,1), filter counts 16/32/64, embedding dimension 64
    Chosen by manual trade-off (Section III-B), not derived from first principles; these determine model capacity and the 1.10 s minimum input length.
  • Acoustic feature parameters = 64 Mel filters, 32 ms Hamming window, 10 ms hop
    Standard choices set in Section IV-A; they define the input representation and hence the model's performance.
assumptions (5)
  • domain assumption Russell's circumplex model with ten emotions mapped to arousal/valence quadrants is a valid target representation.
    The labeling scheme in Fig. 3 and the classification tasks in Table I assume that emotions can be represented as points in a two-dimensional arousal-valence space.
  • domain assumption The sounds produced by a person touching a robot carry sufficient cues to infer the intended gesture and emotional state.
    This is the central hypothesis of the paper; no independent evidence is provided beyond the classification accuracies themselves.
  • domain assumption The split of clips into training, validation, and test sets yields unbiased accuracy estimates.
    Section IV-A gives only per-set sample counts and does not state that participants are kept separate, so the evaluation assumes no leakage from repeated clips of the same participant.
  • domain assumption Pepper's Atom E3845 processor with 11.46 GFLOPs per second defines the real-time deployment constraint.
    Section III-A uses this limit to justify the lightweight design and the real-time claim, but no latency measurement on Pepper hardware is reported.
  • domain assumption Mel-spectrogram features preserve the discriminative information in touch sounds.
    The model uses 64-band Mel features as input; no comparison with alternative front-ends is made, so the sufficiency of this representation is assumed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Sound-Based Recognition of Touch Gestures and Emotions for Enhanced Human-Robot Interaction." pith.science (2026). https://pith.science/paper/67UXD7FD

@misc{pith2026250100038,
  author       = {Pith},
  title        = {Pith review of: Sound-Based Recognition of Touch Gestures and Emotions for Enhanced Human-Robot Interaction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/67UXD7FD}},
  note         = {Machine review of arXiv:2501.00038}
}
read the original abstract

Emotion recognition and touch gesture decoding are crucial for advancing human-robot interaction (HRI), especially in social environments where emotional cues and tactile perception play important roles. However, many humanoid robots, such as Pepper, Nao, and Furhat, lack full-body tactile skin, limiting their ability to engage in touch-based emotional and gesture interactions. In addition, vision-based emotion recognition methods usually face strict GDPR compliance challenges due to the need to collect personal facial data. To address these limitations and avoid privacy issues, this paper studies the potential of using the sounds produced by touching during HRI to recognise tactile gestures and classify emotions along the arousal and valence dimensions. Using a dataset of tactile gestures and emotional interactions from 28 participants with the humanoid robot Pepper, we design an audio-only lightweight touch gesture and emotion recognition model with only 0.24M parameters, 0.94MB model size, and 0.7G FLOPs. Experimental results show that the proposed sound-based touch gesture and emotion recognition model effectively recognises the arousal and valence states of different emotions, as well as various tactile gestures, when the input audio length varies. The proposed model is low-latency and achieves similar results as well-known pretrained audio neural networks (PANNs), but with much smaller FLOPs, parameters, and model size.

Figures

Figures reproduced from arXiv: 2501.00038 by the authors.

Figure 1
Figure 1. The robot Pepper’s physical information1 . please see here1 . In the data collection experiment, participants are asked to express 6 gestures (hold, pat, poke, tickle, tap, rub) and 10 emotions by touching Pepper’s left forearm with spontaneous movements. The sounds produced by these touch movements are recorded by a microphone (the black device in [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The participant interacts with the robot Pepper. [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 4
Figure 4. The proposed lightweight multi-temporal resolution [PITH_FULL_IMAGE:figures/full_fig_p002_4.png] view at source ↗
Figures from the paper (1 more)
Figure 5
Figure 5. Figure 5: Normalized confusion matrix on the test set. [PITH_FULL_IMAGE:figures/full_fig_p004_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 41 canonical work pages

  1. [1]

    Recent advancements in multimodal human–robot interaction,

    H. Su, W. Qi, J. Chen, C. Yang, J. Sandoval, and M. A. Laribi, “Recent advancements in multimodal human–robot interaction,” Frontiers in Neurorobotics, vol. 17, pp. 1084000, 2023

    work page 2023

  2. [2]

    Emotion recognition for human-robot interaction: Recent advances and future perspectives,

    M. Spezialetti, G. Placidi, and S. Rossi, “Emotion recognition for human-robot interaction: Recent advances and future perspectives,” Frontiers in Robotics and AI , vol. 7, pp. 532279, 2020

    work page 2020

  3. [3]

    Affect display recognition through tactile and visual stimuli in a social robot,

    S. Marques-Villarroya, J. J. Gamboa-Montero, C. Jumela-Yedra, J. C. Castillo, and M. A. Salichs, “Affect display recognition through tactile and visual stimuli in a social robot,” in International Conference on Social Robotics . Springer, 2022, pp. 130–140

    work page 2022

  4. [4]

    Humans interacting with multi-robot systems: a natural affect-based approach,

    V . Villani, B. Capelli, C. Secchi, C. Fantuzzi, and L. Sabattini, “Humans interacting with multi-robot systems: a natural affect-based approach,” Autonomous Robots , vol. 44, pp. 601–616, 2020

    work page 2020

  5. [5]

    Gestural behavioral implementation on a humanoid robotic platform for effective social interaction,

    L. Brown and A. M. Howard, “Gestural behavioral implementation on a humanoid robotic platform for effective social interaction,” in IEEE International Symposium on Robot and Human Interactive Communica- tion, 2014, pp. 471–476

    work page 2014

  6. [6]

    A review of surface haptics: Enabling tactile effects on touch surfaces,

    C. Basdogan, F. Giraud, V . Levesque, and S. Choi, “A review of surface haptics: Enabling tactile effects on touch surfaces,” IEEE Transactions on haptics , vol. 13, no. 3, pp. 450–470, 2020

    work page 2020

  7. [7]

    Pepper: The first machine of its kind,

    A. K. Pandey, R. Gelin, and A. Robot, “Pepper: The first machine of its kind,” IEEE Robotics & Automation Magazine , vol. 25, no. 3, pp. 40–48, 2018

    work page 2018

  8. [8]

    Humanoid robot nao: Review of control and motion exploration,

    S. Shamsuddin, L. I. Ismail, H. Yussof, N. I. Zahari, S. Bahari, H. Hashim, and A. Jaffar, “Humanoid robot nao: Review of control and motion exploration,” in IEEE international conference on Control System, Computing and Engineering . IEEE, 2011, pp. 511–516

    work page 2011

Show all 44 references
  1. [9]

    Furhat: a back-projected human-like robot head for multiparty human-machine interaction,

    S. Al Moubayed, J. Beskow, G. Skantze, et al., “Furhat: a back-projected human-like robot head for multiparty human-machine interaction,” in Cognitive Behavioural Systems . Springer, 2012, pp. 114–130

    2012

  2. [10]

    Tactile sens- ing—from humans to humanoids,

    R. S. Dahiya, G. Metta, M. Valle, and G. Sandini, “Tactile sens- ing—from humans to humanoids,” IEEE Transactions on robotics , vol. 26, no. 1, pp. 1–20, 2009

    2009

  3. [11]

    Gelsight: High-resolution robot tactile sensors for estimating geometry and force,

    W. Yuan, S. Dong, and E. H. Adelson, “Gelsight: High-resolution robot tactile sensors for estimating geometry and force,” Sensors, vol. 17, no. 12, pp. 2762, 2017

    2017

  4. [12]

    Tactile sensors: A review,

    M. Meribout, N. A. Takele, O. Derege, N. Rifiki, M. El K., V . Tiwari, and J. Zhong, “Tactile sensors: A review,” Measurement, p. 115332, 2024

    2024

  5. [13]

    Depth camera based hand gesture recognition and its applications in human-computer-interaction,

    Z. Ren, J. Meng, and J. Yuan, “Depth camera based hand gesture recognition and its applications in human-computer-interaction,” in International conference on information, communications & signal processing. IEEE, 2011, pp. 1–5

    2011

  6. [14]

    Vision-based hand gesture recognition for human-robot collaboration: a survey,

    Z. Xia, Q. Lei, Y . Yang, H. Zhang, Y . He, W. Wang, and M. Huang, “Vision-based hand gesture recognition for human-robot collaboration: a survey,” in International Conference on Control, Automation and Robotics. IEEE, 2019, pp. 198–205

    2019

  7. [15]

    Real-time human pose recognition in parts from single depth images,

    J. Shotton, A. Fitzgibbon, M. Cook, T. Sharp, M. Finocchio, R. Moore, A. Kipman, and A. Blake, “Real-time human pose recognition in parts from single depth images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition . IEEE, 2011, pp. 1297–1304

    2011

  8. [16]

    Survey on emotional body gesture recognition,

    F. Noroozi, C. A. Corneanu, D. Kami ´nska, T. Sapi´nski, S. Escalera, and G. Anbarjafari, “Survey on emotional body gesture recognition,” IEEE Transactions on affective computing , vol. 12, no. 2, pp. 505–523, 2018

    2018

  9. [17]

    Vskin: Sensing touch gestures on surfaces of mobile devices using acoustic signals,

    K. Sun, T. Zhao, W. Wang, and L. Xie, “Vskin: Sensing touch gestures on surfaces of mobile devices using acoustic signals,” in Proceedings of Annual International Conference on Mobile Computing and Networking , 2018, pp. 591–605

    2018

  10. [18]

    Sonification of surface tapping changes behavior, surface percep- tion, and emotion,

    A. Tajadura-Jim ´enez, N. Bianchi-Berthouze, E. Furfaro, and F. Bevilac- qua, “Sonification of surface tapping changes behavior, surface percep- tion, and emotion,” IEEE MultiMedia , vol. 22, no. 1, pp. 48–57, 2015

    2015

  11. [19]

    Emotion recognition in human-computer interaction,

    R. Cowie, E. Douglas-Cowie, N. Tsapatsoulis, G. V otsis, S. Kollias, et al., “Emotion recognition in human-computer interaction,” IEEE Signal processing magazine , vol. 18, no. 1, pp. 32–80, 2001

    2001

  12. [20]

    Openear—introducing the munich open-source emotion and affect recognition toolkit,

    F. Eyben, M. W ¨ollmer, and B. Schuller, “Openear—introducing the munich open-source emotion and affect recognition toolkit,” in Interna- tional conference on affective computing and intelligent interaction and workshops. IEEE, 2009, pp. 1–6

    2009

  13. [21]

    A scoping review of the literature on prosodic elements related to emotional speech in human-robot interaction,

    N. Gasteiger, J. Lim, M. Hellou, B. A. MacDonald, and H. S. Ahn, “A scoping review of the literature on prosodic elements related to emotional speech in human-robot interaction,” International Journal of Social Robotics , vol. 16, no. 4, pp. 659–670, 2024

    2024

  14. [22]

    A cross-cultural study of a circumplex model of affect.,

    J. A. Russell, M. Lewicka, and T. Niit, “A cross-cultural study of a circumplex model of affect.,” Journal of personality and social psychology, vol. 57, no. 5, pp. 848, 1989

    1989

  15. [23]

    A circumplex model of affect.,

    J. A. Russell, “A circumplex model of affect.,” Journal of personality and social psychology , vol. 39, no. 6, pp. 1161, 1980

    1980

  16. [24]

    Emotional valence and arousal affect reading in an interactive way: neuroimaging evidence for an approach-withdrawal framework,

    F. M. Citron, M. A. Gray, H. D. Critchley, B. S. Weekes, and E. C. Ferstl, “Emotional valence and arousal affect reading in an interactive way: neuroimaging evidence for an approach-withdrawal framework,” Neuropsychologia, vol. 56, pp. 79–89, 2014

    2014

  17. [25]

    Neural systems subserving valence and arousal during the experience of induced emotions.,

    T. Colibazzi, J. Posner, Z. Wang, D. Gorman, A. Gerber, S. Yu, et al., “Neural systems subserving valence and arousal during the experience of induced emotions.,” Emotion, vol. 10, no. 3, pp. 377, 2010

    2010

  18. [26]

    Understanding urban and natural soundscapes,

    D. Botteldooren, C. Lavandier, A. Preis, D. Dubois, I. Aspuru, et al., “Understanding urban and natural soundscapes,” in F orum Acusticum

  19. [27]

    Soundscape for european cities and landscape: understanding and exchanging,

    D. Botteldooren, T. Andringa, I. Aspuru, L. Brown, D. Dubois, C. Guas- tavino, C. Lavandier, M. Nilsson, and A. Preis, “Soundscape for european cities and landscape: understanding and exchanging,” in COST TD0804 Final conference: Soundscape of European cities and landscapes. S...

    2013

  20. [28]

    Ai-based soundscape analysis: Jointly identifying sound sources and predicting annoyance,

    Y . Hou, Q. Ren, H. Zhang, A. Mitchell, F. Aletta, J. Kang, and D. Botteldooren, “Ai-based soundscape analysis: Jointly identifying sound sources and predicting annoyance,” The Journal of the Acoustical Society of America (JASA) , vol. 154, no. 5, pp. 3145–3157, 2023

    2023

  21. [29]

    Soundscape captioning using sound affective quality network and large language model,

    Y . Hou, Q. Ren, A. Mitchell, W. Wang, J. Kang, T. Belpaeme, and D. Botteldooren, “Soundscape captioning using sound affective quality network and large language model,” arXiv preprint arXiv:2406.05914 , 2024

    2024 arXiv

  22. [30]

    Understanding convolution for semantic segmentation,

    P. Wang, P. Chen, Y . Yuan, D. Liu, Z. Huang, X. Hou, and G. Cottrell, “Understanding convolution for semantic segmentation,” in IEEE winter conference on applications of computer vision . IEEE, 2018, pp. 1451– 1460

    2018

  23. [31]

    Hierarchical shrinkage multiscale network for hyperspectral image classification with hierarchical feature fusion,

    H. Gao, Z. Chen, and C. Li, “Hierarchical shrinkage multiscale network for hyperspectral image classification with hierarchical feature fusion,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , vol. 14, pp. 5760–5772, 2021

    2021

  24. [32]

    Cooperative scene-event modelling for acoustic scene classification,

    Y . Hou, B. Kang, A. Mitchell, W. Wang, J. Kang, and D. Botteldooren, “Cooperative scene-event modelling for acoustic scene classification,” IEEE/ACM Transactions on audio, speech, and language processing , vol. 32, pp. 68–82, 2024

    2024

  25. [33]

    Conveying emotions to robots through touch and sound,

    Q. Ren, R. Proesmans, F. Bossuyt, J. Vanfleteren, F. Wyffels, and T. Belpaeme, “Conveying emotions to robots through touch and sound,” arXiv preprint arXiv:2412.03300 , 2024

    2024 arXiv

  26. [34]

    PANNs: Large-scale pretrained audio neural networks for audio pattern recognition,

    Q. Kong, Y . Cao, T. Iqbal, Y . Wang, W. Wang, and M. D. Plumbley, “PANNs: Large-scale pretrained audio neural networks for audio pattern recognition,” IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 28, pp. 2880–2894, 2020

    2020

  27. [35]

    Dropout: a simple way to prevent neural networks from overfitting,

    N. Srivastava, G. Hinton, A. Krizhevsky, I. Sutskever, et al., “Dropout: a simple way to prevent neural networks from overfitting,” Journal of Machine Learning Research , vol. 15, no. 1, pp. 1929–1958, 2014

    1929

  28. [36]

    Adam: A method for stochastic optimization,

    D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” in Proc. of ICLR , 2015

    2015

  29. [37]

    Emotion and perception: The role of affective information,

    J. R. Zadra and G. L. Clore, “Emotion and perception: The role of affective information,” Wiley interdisciplinary reviews: cognitive science, vol. 2, no. 6, pp. 676–685, 2011

    2011

  30. [38]

    The effect of context on choice and value,

    B. De Martino, “The effect of context on choice and value,” in Neuroscience of Preference and Choice , pp. 93–119. Elsevier, 2012

    2012

  31. [39]

    Comparisons of various types of normality tests,

    B. W. Yap and C. H. Sim, “Comparisons of various types of normality tests,” Journal of Statistical Computation and Simulation , vol. 81, no. 12, pp. 2141–2155, 2011

    2011

  32. [40]

    Con- tinuous variable analyses: t-test, mann–whitney, wilcoxin rank,

    M. D. Riina, C. Stambaugh, N. Stambaugh, and K. E. Huber, “Con- tinuous variable analyses: t-test, mann–whitney, wilcoxin rank,” in Translational Radiation Oncology , pp. 153–163. Elsevier, 2023

    2023

  33. [41]

    On rhythmic and discrete movements: reflections, definitions and implications for motor control,

    N. Hogan and D. Sternad, “On rhythmic and discrete movements: reflections, definitions and implications for motor control,” Experimental brain research, vol. 181, pp. 13–30, 2007

    2007

  34. [42]

    Mo- bilenetv2: Inverted residuals and linear bottlenecks,

    M. Sandler, A. Howard, M. Zhu, A. Zhmoginov, and L. Chen, “Mo- bilenetv2: Inverted residuals and linear bottlenecks,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2018, pp. 4510–4520

    2018

  35. [43]

    AudioSet: An ontology and human-labeled dataset for audio events,

    J. Gemmeke, D. Ellis, D. Freedman, A. Jansen, W. Lawrence, R. C. Moore, et al., “AudioSet: An ontology and human-labeled dataset for audio events,” in IEEE international conference on acoustics, speech and signal processing (ICASSP) . IEEE, 2017, pp. 776–780

    2017

  36. [2011]

    2047–2052

    European Accoustics Association, 2011, pp. 2047–2052

    2011

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.