Cross-species Behavior Analysis with Attention- based Domain-adversarial Deep Neural Networks Takuya Maekawa (
[email protected] ) Osaka University https://orcid.org/0000-0002-7227-580X Daiki Higashide Osaka University Takahiro Hara Osaka University Kentarou Matsumura Okayama University Kaoru Ide Doshisha University Takahisa Miyatake Okayama University https://orcid.org/0000-0002-5476-0676 Koutarou Kimura Nagoya City University https://orcid.org/0000-0002-3359-1578 Susumu Takahashi Doshisha University Article Keywords: behavior analysis, deep neural networks, cross-species features Posted Date: December 15th, 2020 DOI: https://doi.org/10.21203/rs.3.rs-123107/v1 License: This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License 1 Cross-species Behavior Analysis with Attention-based 2 Domain-adversarial Deep Neural Networks 1,∗ 1 1 2 3 3 Takuya Maekawa , Daiki Higashide , Takahiro Hara , Kentarou Matsumura , Kaoru Ide , Takahisa 2 4 3 4 Miyatake , Koutarou D. Kimura & Susumu Takahashi 1 5 Graduate School of Information Science and Technology, Osaka University, Osaka 565-0871, 6 Japan 2 7 Graduate School of Environmental and Life Science, Okayama University, Okayama 700-8530, 8 Japan 3 9 Graduate School of Brain Science, Doshisha University, Kyoto 610-0321, Japan 4 10 Graduate School of Natural Sciences, Nagoya City University, Aichi 467-8501 Japan 11 *Correspondence to Takuya Maekawa. 12 Since the variables inherent to various diseases cannot be controlled directly in humans, 13 behavioral dysfunctions have been examined in model organisms, leading to better under- 14 standing their underlying mechanisms. However, because the spatial and temporal scales 15 of animal locomotion vary widely among species, conventional statistical analyses cannot be 16 used to discover knowledge from the locomotion data.