Cited by 7Tactile and Sensory InteractionsEEG and Brain-Computer InterfacesMultisensory perception and integration

Autoencoding sensory substitution

Open access
Tóth, Viktor, Parkkonen, Lauri·2019·Unpublished

Abstract

Tens of millions of people live blind, and their number is ever increasing. Visual-to-auditory sensory substitution (SS) encompasses a family of cheap, generic solutions to assist the visually impaired by conveying visual information through sound. The required SS training is lengthy: months of effort is necessary to reach a practical level of adaptation. There are two reasons for the tedious training process: the elongated substituting audio signal, and the disregard for the compressive characteristics of the human hearing system. To overcome these obstacles, we developed a novel class of SS methods, by training deep recurrent autoencoders for image-to-sound conversion. We successfully trained deep learning models on different datasets to execute visual-to-auditory stimulus conversion. By constraining the visual space, we demonstrated the viability of shortened substituting audio signals, while proposing mechanisms, such as the integration of computational hearing models, to optimally convey visual features in the substituting stimulus as perceptually discernible auditory components. We tested our approach in two separate cases. In the first experiment, the author went blindfolded for 5 days, while performing SS training on hand posture discrimination. The second experiment assessed the accuracy of reaching movements towards objects on a table. In both test cases, above-chance-level accuracy was attained after a few hours of training. Our novel SS architecture broadens the horizon of rehabilitation methods engineered for the visually impaired. Further improvements on the proposed model shall yield hastened rehabilitation of the blind and a wider adaptation of SS devices as a consequence.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

DataCite · 10.13140/rg.2.2.10576.87048

Cites

158 of 161 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • AEV2A makes it difficult for a participant to understand the conversion scheme because there is no consistent correlation between changes in input image and resulting sound.

    “However, this method makes it difficult for a participant to understand the conversion scheme because there is no consistent correlation between changes in the input image and the resulting sound obtained from AEV2A.”

    From Deep Learning-Based Optimization of Visual–Auditory Sensory Substitution (2023) · cites this paper as [23]

    Section 2.5.1 (page unavailable) says implicit conversion acts as a black box and that explicit methods are easier to grasp, supporting the difficulty of understanding AEV2A’s conversion scheme. It does not support the asserted absence of a consistent image–sound correlation: Section 4.1 reports measured correlations between visual and audio features, including r = .97 and r = .99. (LLM, conf=0.87)

    Experimental — accuracy not yet measured

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

1 of the 162 references couldn’t be matched to a known paper.

Research works printed but not linked: 25

  1. [2][2] B. Roska and J.-A. Sahel, 'Restoring vision,' Nature , vol. 557, no. 7705, pp. 359-367, 2018. -
  2. [37][37] W. Gerstner and W. M. Kistler, Spiking neuron models: Single neurons, populations, plasticity . Cambridge university press, 2002. -
  3. [38][38] J. Hawkins, S. Ahmad, and Y. Cui, 'A Theory of How Columns in the Neocortex Enable Learning the Structure of the World,' Frontiers in Neural Circuits , vol. 11, 2017. [Online]. Available: https://www.frontiersin.org/ articles/10.3389/fncir.2017.00081/full - · DOI
  4. [42][42] L. Renier, A. G. De Volder, and J. P. Rauschecker, 'Cortical plasticity and preserved function in early blindness,' Neuroscience & Biobehavioral Reviews , vol. 41, pp. 53-63, Apr. 2014. [Online]. Available: http://www.sciencedirect.com/science/article/pii/S0149763413000328 -
  5. [49][49] B. Richmond, 'Information Coding,' Science , vol. 294, no. 5551, pp. 2493-2494, Dec. 2001. [Online]. Available: http://science.sciencemag.org/ content/294/5551/2493 -
  6. [86][86] J. Rodríguez, S. T. Neely, H. Patra, J. Kopun, W. Jesteadt, H. Tan, and M. P. Gorga, 'The role of suppression in psychophysical toneon-tone masking,' The Journal of the Acoustical Society of America , vol. 127, no. 1, pp. 361-369, Jan. 2010. [Online]. Available: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2821167/ -
  7. [96][96] N. Fujiki, K. A. J. Riederer, V. Jousmäki, J. P. Mäkelä, and R. Hari, 'Human cortical representation of virtual auditory space: differences between sound azimuth and elevation,' The European Journal of Neuroscience , vol. 16, no. 11, pp. 2207-2213, Dec. 2002. -
  8. [122][122] C. Graulty, O. Papaioannou, P. Bauer, M. A. Pitts, and E. Canseco-Gonzalez, 'Hearing Shapes: Event-related Potentials Reveal the Time Course of AuditoryVisual Sensory Substitution,' Journal of Cognitive Neuroscience , vol. 30, no. 4, pp. 498-513, 2018. -
  9. [128][128] L. G. Cohen, R. A. Weeks, N. Sadato, P. Celnik, K. Ishii, and M. Hallett, 'Period of susceptibility for cross-modal plasticity in the blind,' Annals of Neurology , vol. 45, no. 4, pp. 451-460, Apr. 1999. [Online]. Available: https://onlinelibrary.wiley.com/doi/abs/10.1002/1531-8249%28199904% 2945%3A4%3C451%3A%3AAID-ANA6%3E3.0.CO%3B2-B - · DOI
  10. [134][134] P. Voss, M. Lassonde, F. Gougoux, M. Fortin, J.-P. Guillemot, and F. Lepore, 'Early- and late-onset blind individuals show supra-normal auditory abilities in far-space,' Current biology: CB , vol. 14, no. 19, pp. 1734-1738, Oct. 2004. -
  11. [145][145] P. Bach-Y-Rita, C. C. Collins, F. A. Saunders, B. White, and L. Scadden, 'Vision Substitution by Tactile Image Projection,' Nature , vol. 221, no. 5184, pp. 963-964, Mar. 1969. [Online]. Available: https://www.nature.com/articles/221963a0 -
  12. [153][153] M. Bujacz, K. Kropidlowski, G. Ivanica, A. Moldoveanu, C. Saitis, A. Csapo, G. Wersenyi, S. Spagnol, O. I. Johannesson, R. Unnthorsson, M. Rotnicki, and P. Witek, 'Sound of Vision - Spatial Audio Output and Sonification Approaches,' in Computers Helping People with Special Needs , ser. Lecture Notes in Computer Science, K. Miesenberger, C. Bühler, and P. Penaz, Eds. Springer International Publishing, 2016, pp. 202-209. -
  13. [155][155] G. V. Elli, S. Benetti, and O. Collignon, 'Is there a future for sensory substitution outside academic laboratories?' Multisensory Research , vol. 27, no. 5-6, pp. 271-291, 2014. -
  14. [156][156] M. A. De Niear, J.-P. Noel, and M. T. Wallace, 'The Impact of Feedback on the Different Time Courses of Multisensory Temporal Recalibration,' Neural Plasticity , vol. 2017, 2017. [Online]. Available: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5339631/ -
  15. [161][161] B. Durette, N. Louveton, D. Alleysson, and J. Hérault, 'Visuo-auditory sensory substitution for mobility assistance: testing TheVIBE,' in Workshop on Computer Vision Applications for the Visually Impaired . Marseille, France: James Coughlan and Roberto Manduchi, Oct. 2008. [Online]. Available: https://hal.inria.fr/inria-00325414 -
  16. [167][167] T. C. Kietzmann, P. McClure, and N. Kriegeskorte, 'Deep Neural Networks In Computational Neuroscience,' bioRxiv , p. 133504, May 2017. [Online]. Available: https://www.biorxiv.org/content/early/2017/05/04/133504 -
  17. [170][170] D. L. K. Yamins, H. Hong, C. F. Cadieu, E. A. Solomon, D. Seibert, and J. J. DiCarlo, 'Performance-optimized hierarchical models predict neural responses in higher visual cortex,' Proceedings of the National Academy of Sciences , vol. 111, no. 23, pp. 8619-8624, Jun. 2014. [Online]. Available: http://www.pnas.org/content/111/23/8619 -
  18. [172][172] A. Banino, C. Barry, B. Uria, C. Blundell, T. Lillicrap, P. Mirowski, A. Pritzel, M. J. Chadwick, T. Degris, J. Modayil, G. Wayne, H. Soyer, F. Viola, B. Zhang, R. Goroshin, N. Rabinowitz, R. Pascanu, C. Beattie, S. Petersen, <!-- image --> A. Sadik, S. Gaffney, H. King, K. Kavukcuoglu, D. Hassabis, R. Hadsell, and D. Kumaran, 'Vector-based navigation using grid-like representations in artificial agents,' Nature , vol. 557, no. 7705, p. 429, May 2018. [Online]. Available: https://www.nature.com/articles/s41586-018-0102-6 -
  19. [175][175] I. Higgins, L. Matthey, X. Glorot, A. Pal, B. Uria, C. Blundell, S. Mohamed, and A. Lerchner, 'Early Visual Concept Learning with Unsupervised Deep Learning,' arXiv:1606.05579 [cs, q-bio, stat] , Jun. 2016, arXiv: 1606.05579. [Online]. Available: http://arxiv.org/abs/1606.05579 - · arXiv
  20. [180][180] A. Makhzani, J. Shlens, N. Jaitly, I. Goodfellow, and B. Frey, 'Adversarial Autoencoders,' arXiv:1511.05644 [cs] , Nov. 2015, arXiv: 1511.05644. [Online]. Available: http://arxiv.org/abs/1511.05644 - · arXiv
  21. [183][183] A. v. d. Oord, Y. Li, I. Babuschkin, K. Simonyan, O. Vinyals, K. Kavukcuoglu, G. v. d. Driessche, E. Lockhart, L. C. Cobo, F. Stimberg, N. Casagrande, D. Grewe, S. Noury, S. Dieleman, E. Elsen, N. Kalchbrenner, H. Zen, A. Graves, H. King, T. Walters, D. Belov, and D. Hassabis, 'Parallel WaveNet: Fast High-Fidelity Speech Synthesis,' arXiv:1711.10433 [cs] , Nov. 2017, arXiv: 1711.10433. [Online]. Available: http://arxiv.org/abs/1711.10433 - · arXiv
  22. [188][188] R. Pascanu, T. Mikolov, and Y. Bengio, 'On the difficulty of training Recurrent Neural Networks,' Nov. 2012. [Online]. Available: https://arxiv.org/abs/1211.5063 - · arXiv
  23. [189][189] S. Santurkar, D. Tsipras, A. Ilyas, and A. Madry, 'How Does Batch Normalization Help Optimization?' May 2018. [Online]. Available: https://arxiv.org/abs/1805.11604 - · arXiv
  24. [192][192] Z. Liu, P. Luo, X. Wang, and X. Tang, 'Deep learning face attributes in the wild,' in Proceedings of International Conference on Computer Vision (ICCV) , 2015. -
  25. [194][194] C. K. Sønderby, T. Raiko, L. Maaløe, S. K. Sønderby, and O. Winther, 'Ladder Variational Autoencoders,' arXiv:1602.02282 [cs, stat] , Feb. 2016, arXiv: 1602.02282. [Online]. Available: http://arxiv.org/abs/1602.02282 - · arXiv

Cited by

Sorted by citations

Authors
Tóth, V., & Parkkonen, L.
Corresponding author
Tóth, Viktor
Venue
Unpublished
Publisher
Unpublished
Published
2019
ISSN
2331-8422 (ISSN-L)
OpenAlex
W2951582018
Cited by
7
References
162
Citations by year
2025: 1 · 2023: 1 · 2022: 3 · 2021: 1 · 2020: 1
Type
posted content
Language
en
Open access
Locations
  • arXiv (Cornell University) (submittedVersion)
  • Aaltodoc (Aalto University) (submittedVersion)
  • arXiv (Cornell University)
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Tactile and Sensory Interactions
  • Life Sciences › Neuroscience › Cognitive Neuroscience › EEG and Brain-Computer Interfaces
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Multisensory perception and integration
Keywords
sensory substitution, recurrent autoencoder, blind rehabilitation, auditory display, reaching movements, audio compression

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review