Illusory Motion Reproduced by Deep Neural Networks Trained for Prediction
The cerebral cortex predicts visual motion to adapt human behavior to surrounding objects moving in real time. Although the underlying mechanisms are still unknown, predictive coding is one of the leading theories. Predictive coding assumes that the brain's internal models (which are acquired t...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2018-03-01
|
| Series: | Frontiers in Psychology |
| Subjects: | |
| Online Access: | http://journal.frontiersin.org/article/10.3389/fpsyg.2018.00345/full |
| _version_ | 1819129105933664256 |
|---|---|
| author | Eiji Watanabe Eiji Watanabe Akiyoshi Kitaoka Kiwako Sakamoto Kiwako Sakamoto Masaki Yasugi Kenta Tanaka |
| author_facet | Eiji Watanabe Eiji Watanabe Akiyoshi Kitaoka Kiwako Sakamoto Kiwako Sakamoto Masaki Yasugi Kenta Tanaka |
| author_sort | Eiji Watanabe |
| collection | DOAJ |
| description | The cerebral cortex predicts visual motion to adapt human behavior to surrounding objects moving in real time. Although the underlying mechanisms are still unknown, predictive coding is one of the leading theories. Predictive coding assumes that the brain's internal models (which are acquired through learning) predict the visual world at all times and that errors between the prediction and the actual sensory input further refine the internal models. In the past year, deep neural networks based on predictive coding were reported for a video prediction machine called PredNet. If the theory substantially reproduces the visual information processing of the cerebral cortex, then PredNet can be expected to represent the human visual perception of motion. In this study, PredNet was trained with natural scene videos of the self-motion of the viewer, and the motion prediction ability of the obtained computer model was verified using unlearned videos. We found that the computer model accurately predicted the magnitude and direction of motion of a rotating propeller in unlearned videos. Surprisingly, it also represented the rotational motion for illusion images that were not moving physically, much like human visual perception. While the trained network accurately reproduced the direction of illusory rotation, it did not detect motion components in negative control pictures wherein people do not perceive illusory motion. This research supports the exciting idea that the mechanism assumed by the predictive coding theory is one of basis of motion illusion generation. Using sensory illusions as indicators of human perception, deep neural networks are expected to contribute significantly to the development of brain research. |
| first_indexed | 2024-12-22T08:38:26Z |
| format | Article |
| id | doaj.art-1d1753ef96a2421c8b3c7876e542b181 |
| institution | Directory Open Access Journal |
| issn | 1664-1078 |
| language | English |
| last_indexed | 2024-12-22T08:38:26Z |
| publishDate | 2018-03-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Psychology |
| spelling | doaj.art-1d1753ef96a2421c8b3c7876e542b1812022-12-21T18:32:17ZengFrontiers Media S.A.Frontiers in Psychology1664-10782018-03-01910.3389/fpsyg.2018.00345340023Illusory Motion Reproduced by Deep Neural Networks Trained for PredictionEiji Watanabe0Eiji Watanabe1Akiyoshi Kitaoka2Kiwako Sakamoto3Kiwako Sakamoto4Masaki Yasugi5Kenta Tanaka6Laboratory of Neurophysiology, National Institute for Basic Biology, Okazaki, JapanDepartment of Basic Biology, The Graduate University for Advanced Studies (SOKENDAI), Miura, JapanDepartment of Psychology, Ritsumeikan University, Kyoto, JapanDepartment of Physiological Sciences, The Graduate University for Advanced Studies (SOKENDAI), Miura, JapanDivision of Integrative Physiology, National Institute for Physiological Sciences (NIPS), Okazaki, JapanLaboratory of Neurophysiology, National Institute for Basic Biology, Okazaki, JapanSakura Research Office, Wako, JapanThe cerebral cortex predicts visual motion to adapt human behavior to surrounding objects moving in real time. Although the underlying mechanisms are still unknown, predictive coding is one of the leading theories. Predictive coding assumes that the brain's internal models (which are acquired through learning) predict the visual world at all times and that errors between the prediction and the actual sensory input further refine the internal models. In the past year, deep neural networks based on predictive coding were reported for a video prediction machine called PredNet. If the theory substantially reproduces the visual information processing of the cerebral cortex, then PredNet can be expected to represent the human visual perception of motion. In this study, PredNet was trained with natural scene videos of the self-motion of the viewer, and the motion prediction ability of the obtained computer model was verified using unlearned videos. We found that the computer model accurately predicted the magnitude and direction of motion of a rotating propeller in unlearned videos. Surprisingly, it also represented the rotational motion for illusion images that were not moving physically, much like human visual perception. While the trained network accurately reproduced the direction of illusory rotation, it did not detect motion components in negative control pictures wherein people do not perceive illusory motion. This research supports the exciting idea that the mechanism assumed by the predictive coding theory is one of basis of motion illusion generation. Using sensory illusions as indicators of human perception, deep neural networks are expected to contribute significantly to the development of brain research.http://journal.frontiersin.org/article/10.3389/fpsyg.2018.00345/fullvisual illusionspredictive codingdeep learningartificial intelligencecerebral cortex |
| spellingShingle | Eiji Watanabe Eiji Watanabe Akiyoshi Kitaoka Kiwako Sakamoto Kiwako Sakamoto Masaki Yasugi Kenta Tanaka Illusory Motion Reproduced by Deep Neural Networks Trained for Prediction Frontiers in Psychology visual illusions predictive coding deep learning artificial intelligence cerebral cortex |
| title | Illusory Motion Reproduced by Deep Neural Networks Trained for Prediction |
| title_full | Illusory Motion Reproduced by Deep Neural Networks Trained for Prediction |
| title_fullStr | Illusory Motion Reproduced by Deep Neural Networks Trained for Prediction |
| title_full_unstemmed | Illusory Motion Reproduced by Deep Neural Networks Trained for Prediction |
| title_short | Illusory Motion Reproduced by Deep Neural Networks Trained for Prediction |
| title_sort | illusory motion reproduced by deep neural networks trained for prediction |
| topic | visual illusions predictive coding deep learning artificial intelligence cerebral cortex |
| url | http://journal.frontiersin.org/article/10.3389/fpsyg.2018.00345/full |
| work_keys_str_mv | AT eijiwatanabe illusorymotionreproducedbydeepneuralnetworkstrainedforprediction AT eijiwatanabe illusorymotionreproducedbydeepneuralnetworkstrainedforprediction AT akiyoshikitaoka illusorymotionreproducedbydeepneuralnetworkstrainedforprediction AT kiwakosakamoto illusorymotionreproducedbydeepneuralnetworkstrainedforprediction AT kiwakosakamoto illusorymotionreproducedbydeepneuralnetworkstrainedforprediction AT masakiyasugi illusorymotionreproducedbydeepneuralnetworkstrainedforprediction AT kentatanaka illusorymotionreproducedbydeepneuralnetworkstrainedforprediction |