Updated on 2026/11/02

写真a

 
YAMASHITA Wakayo
 
Organization
Research Field in Engineering, Science and Engineering Area Graduate School of Science and Engineering (Engineering) Department of Informatics Informatics Program Associate Professor
Title
Associate Professor
Degree
(2009.3 Kagoshima University)
(2006.3 Kagoshima University)
(2004.3 Kagoshima University)

Research Interests

  • 認知科学

  • 乳幼児

  • 発達

  • 物体認識

Research Areas

Cognitive science, 視覚神経科学, 発達

Research History

  • 2022.4    Kagoshima University   Associate Professor

  • 2020.4    Kagoshima University   Research Field in Engineering, Science and Engineering Area Graduate School of Science and Engineering (Engineering) Department of Engineering Information Science and Biomedical Engineering Program   Assistant Professor

  • 2013.4 - 2020.3    Kagoshima University   Research Field in Engineering, Science and Engineering Area Graduate School of Science and Engineering (Engineering) Information Science and Biomedical Engineering Course   Assistant Professor

Professional Memberships

  • 2015.10    日本神経科学学会

  • 2015.10    北米神経科学学会

  • 2015.10    日本視覚学会

  • 2015.10    Vision Science Society

 

Papers

  • SUZUKI Wataru, ASAKAWA Shinichi, YAMASHITA Wakayo, TAKEICHI Hiroshige .  Representation and Synthesis of Velocity Vector Fields in Latent Space Using VAE-GP .  Proceedings of the Annual Conference of JSAIJSAI2026 ( 0 ) 2G1OS2104 - 2G1OS2104   2026

     More details

    Language:Japanese   Publisher:The Japanese Society for Artificial Intelligence  

    <p>This study uses machine learning to model and interpret visual motion, comparing and contrasting it with the way humans perceive fluid viscosity from optical flow. Optical flow represents two-dimensional motion on the retina that arises from the three-dimensional motion of objects and observers. Conventional methods face challenges such as ambiguity, occlusion and dimensionality reduction. Although recent FlowNet-based approaches have improved performance, they lack interpretability. This study investigates the application of VAE-GPLVM and VAE-GP, which integrate deep learning with Gaussian processes (GPs). Filtered visual motion data were compressed into a latent space using VAE. The former further compressed the VAE using GPLVM, while the latter constructed the latent space with VAE and incorporated GP as the prior distribution. When trained on fluid videos with varying viscosities that had been converted into spatiotemporal vector fields, only VAE-GP demonstrated superior reconstruction performance in terms of mean squared error (MSE) and structural similarity indices, as well as effective interpolation and extrapolation within the latent space. These results suggest that GP modelling with smooth latent manifolds has a suitable property that is congruent with human natural perception of movies.</p>

    DOI: 10.11517/pjsai.jsai2026.0_2g1os2104

    CiNii Research

  • Takeichi H, Suzuki W, Yamashita W, Hiyama A. .  Perception of nonrigid structures from motion using tracking image gradient vectors. .  Frontiers in Psychology16   1 - 10   2025.8Reviewed

     More details

    Language:English   Publishing type:Research paper (scientific journal)   Publisher:Frontiers Media SA  

    Introduction

    A subset of the true optical flow can be extracted by constructing a vector field that represents image gradients and then tracking vectors in this vector field. This pseudo-flow (p-flow) subset effectively visualizes nonrigid motion and leads to the perception of nonrigid structure from motion. In this study, we investigate whether the human sensory system can extract information about the physical properties of inanimate fluid, especially viscosity, from the p-flow.

    Methods

    Computer-generated movies of flowing liquid were constructed using the p-flow algorithm and the Lucas–Kanade method. The movies featured liquids of different viscosities in the form of point-light displays. The viscosity of the fluid in various subsets of these movies was then estimated by 312 participants.

    Results

    The error, i.e., difference between expected and actual ratings showed smaller variability across repeated trials and the mean response time was significantly shorter when using the p-flow than with the conventional Lucas–Kanade method.

    Discussion

    Our results suggest that the p-flow enables a more reliable viscosity rating, which could be related to the local constraint used in the algorithm.

    DOI: 10.3389/fpsyg.2025.1586648

    URL: https://www.frontiersin.org/articles/10.3389/fpsyg.2025.1586648/full

    Scopus

    PubMed

  • Sung-En Chien, Su-Ling Yeh, Wakayo Yamashita, Sei-Ichi Tsujimura .  Enhanced human contrast sensitivity with increased stimulation of melanopsin in intrinsically photosensitive retinal ganglion cells. .  Vision research209   108271 - 108271   2023.6International journal

     More details

    Language:English   Publishing type:Research paper (scientific journal)  

    The intrinsically photosensitive retinal ganglion cells (ipRGCs) are known to serve non-image-forming functions, such as photoentrainment of the circadian rhythm and pupillary light reflex. However, how they affect human spatial vision is largely unknown. The spatial contrast sensitivity function (CSF), which measures contrast sensitivity as a function of spatial frequency, was used in the current study to investigate the function of ipRGCs in pattern vision. To compare the effects of different background lights on the CSF, we utilized the silent substitution technique. We manipulated the stimulation level of melanopsin (i.e., the visual pigment of ipRGCs) from the background light while keeping the cone stimulations constant, or vice versa. We conducted four experiments to measure the CSFs at various spatial frequencies, eccentricities, and levels of background luminance. Results showed that melanopsin stimulation from the background light enhances spatial contrast sensitivity across different eccentricities and luminance levels. Our finding that melanopsin contributes to CSF, combined with the receptive field analysis, suggests a role for the magnocellular pathway and challenges the conventional view that ipRGCs are primarily responsible for non-visual functions.

    DOI: 10.1016/j.visres.2023.108271

    URL: https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85162111694&origin=inward

    Scopus

    PubMed

  • Tsujimura S.I., Matsumoto A., Yamashita W. .  Intrinsic Phase Difference between Cone and Melanopsin Signals in the Pupillary Pathway .  Proceedings - 2023 24th International Conference on Control Systems and Computer Science, CSCS 2023   253 - 260   2023

     More details

    Language:Japanese   Publisher:Proceedings - 2023 24th International Conference on Control Systems and Computer Science, CSCS 2023  

    Humans have adapted to a light environment with natural light in the course of evolution. A novel light environment control system is needed to address the problems of inadequate adaptation to the modern artificial light environment. For a long time, only cone and rod cells were thought to be photoreceptors in the retina (i.e. light sensors in human), but a new photoreceptor was discovered around 2000. These photoreceptors are called melanopsin ganglion cells (ipRGCs: intrinsically photoreceptive retinal ganglion cells). However, at present, the understanding of the functions of melanopsin cells is extremely limited and little is known. For example, many people are getting more sunlight in the morning to improve their sleep quality, wearing blue light-cutting glasses or using the Night Shift function in iOS, suggesting that it is important to activate or inhibit melanopsin cells depending on the light environment, but it is necessary to understand the function of melanopsin cells in order to make appropriate decisions on when to activate or inhibit melanopsin cells. Melanopsin cells have a slower response than cones. They receive signals from classical photoreceptors, indicating that cone and melanopsin signals are integrated at the retina. In the present study, we measured the intrinsic phase difference between cone and melanopsin signals in the pupillary pathway using a silent-substitution technique. The goal of the present study was to investigate how these signals are temporally integrated. We used three different test stimuli: (i) varying melanopsin stimulation without changing L-, M-, and S-cone stimulation (melanopsin stimulus); (ii) varying L-, M-, and Scone stimulation only, without changing melanopsin stimulation (cone stimulus); and (iii) varying the radiant flux of the stimuli, without changing the spectral composition, which reduces or increases the radiant flux uniformly at all wavelengths (lightflux stimulus). Consistent with previous studies, we found a delayed pupillary response to the melanopsin stimulus. The cone signal leads the melanopsin signal by approximately 100 ms in onset constriction of the pupil response. In addition, we measured pupillary responses to the light-flux stimulus, consisting of cone and melanopsin stimuli. The timings of the melanopsin stimuli were set at variable physical phases. When we estimated the intrinsic phase difference from the first harmonic component of the pupil trace, the melanopsin signal leads the cone signal by approximately 32° and the contribution of cone signals is approximately five times greater than that of melanopsin signals at a temporal frequency of 0.5 Hz; this could be accounted for by a linear summation model of cone and melanopsin signals. These results suggest that the intrinsic phase difference between the cone and melanopsin signals is not a simple latency difference between photoreceptors at the retina. The difference in onset time of pupil constriction could be explained by the simple latency difference between cone and melanopsin photoreception, whereas the intrinsic phase difference could be explained by an integration process in the melanopsin receptive field.

    DOI: 10.1109/CSCS59211.2023.00047

    URL: https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85170260152&origin=inward

    Scopus

  • Suzuki W, Hiyama A, Ichinohe N, Yamashita W, Seno T & Takeichi H. .  Visualization by P-flow: gradient- and feature-based optical flow and vector fields extracted from image analysis. .  Journal of the Optical Society of America A37 ( 12 ) 1958 - 1964   2020.12Reviewed

     More details

    Language:English   Publishing type:Research paper (scientific journal)  

  • Chien, SE., Chen, YC., Matsumoto, A., Yamashita, W., Shih, KT., Tsujimura, S., Yeh,SL. .  The modulation of background color on perceiving audiovisual simultaneity .  Vision Research172   2020.6Reviewed

     More details

    Language:English   Publishing type:Research paper (scientific journal)  

    DOI: 10.1016/j.visres.2020.04.009

  • 山下和香代,林潤一,竹原慎太郎,王鋼 .  観察角度をまたいだ物体弁別の学習に対する内挿角画像呈示の促進作用. .  生体医工学46 ( 5 ) 514 - 521   2008.10Reviewed

     More details

    Language:Japanese   Publishing type:Research paper (scientific journal)  

▼display all

Presentations

  • 坂本 雅仁, 松崎 圭祐, 山下 和香代, 辻村 誠一   瞳孔反応の測定による錐体細胞とメラノプシン細胞間の潜時差の推定  

    Vision  2017.1  日本視覚学会

  • 松元 明子, 山下 和香代, 辻村 誠一   錐体細胞とメラノプシン細胞へのON及びOFF刺激による明るさ知覚への影響  

    Vision  2018.1  日本視覚学会

  • Wang Gang, Obama Shinji, Yamashita Wakayo, Sugihara Tadashi, Tanaka Keiji   視力-不変識別は個人視力の経験に依存する(View-invariant discrimination depends on the experience of individual views)  

    Neuroscience Research  2005.7  エルゼビア・ジャパン(株)

     More details

    Language:English  

  • Yamashita Wakayo, Hayashi Junichi, Murakami Tomoki, Wang Gang   視界関連学習は挿入された視界の経験に依存する(View association learning depends on the experience of interpolated views)  

    Neuroscience Research  2006.7  エルゼビア・ジャパン(株)

     More details

    Language:English  

  • 山下 和香代, 金沢 創, 山口 真美   3次元物体としての顔認識に対する視線の効果の発達的検討  

    Vision  2011.1  日本視覚学会

     More details

    Language:Japanese  

Research Projects

  • 映像情報の視聴に関する子どもの知覚認知機能発達の解明

    Grant number:24K05791  2024.4 - 2027.3

    日本学術振興会  科学研究費補助金  基盤研究(C)

    山下 和香代

      More details

    Authorship:Principal investigator 

    Grant amount:\4680000 ( Direct Cost: \3600000 、 Indirect Cost:\1080000 )

  • 視覚的自然場面における人物の動き処理に関する発達過程の解明

    2021.4 - 2024.3

    科学研究費補助金  基盤研究(C)

      More details

    Authorship:Principal investigator  Grant type:Competitive

  • 視覚的自然場面における人物の動き処理に関する発達過程の解明

    Grant number:21K12609  2021.4 - 2024.3

    日本学術振興会  科学研究費助成事業  基盤研究(C)

    山下 和香代

      More details

    Grant amount:\4160000 ( Direct Cost: \3200000 、 Indirect Cost:\960000 )

    本研究は、自然場面において動きだけから人物検出および人物同定といったヒトにおける知覚認知機能の発達の解明を目的とする。動きだけから人物検出及び人物同定といった知覚認知が可能かどうかを検証するために、様々な事象、事物などの情報が重畳した日常環境(自然場面)動画を取得し、その自然場面から抽出されたオプティカルフローに沿ってランダムに分布するドットが動く動画(ドット動画)を用意した。本年度は、成人に対して、自然場面動画とそのドット動画を用い人物検出課題を実施した。成人は、自然場面動画と同じようにドット動画においても人物検出率が高かった。一方、自然場面動画とそのドット動画を倒立提示すると、自然場面動画では人物検出率が高かったが、ドット動画では人物検出率が低下した。このことから、動きだけから人物を検出することが可能であること、そして、動きだけから人物を検出することに倒立効果が生じることが示唆された。また、実験では人物検出課題遂行中の成人の眼球運動を計測しており、データを分析中である。2022年度は、さらに眼球運動計測のデータ分析を進め、自然場面動画観察時とドット動画観察時による比較や正立提示観察時と倒立提示観察時の比較など行い、視線と動きだけから人物を検出することの関係を明らかにする予定である。これら一連の分析結果については、研究成果として学会発表や学術論文として公表することを目指す。

  • Irradiance encoding process in the nonimage-forming pathway

    Grant number:17H01808  2017.4 - 2021.3

    Japan Society for the Promotion of Science  Grants-in-Aid for Scientific Research Grant-in-Aid for Scientific Research (B)  Grant-in-Aid for Scientific Research (B)

    Tsujimura Sei-ichi

      More details

    Grant amount:\17420000 ( Direct Cost: \13400000 、 Indirect Cost:\4020000 )

    There are photoreceptors in the retina, cone and rod cells. In addition, the third type of photoreceptor was discovered around 2000 which is called melanopsin cells. In this study, we aimed to clarify how the non-imaging pathway originating from melanopsin cells processes and sends light signals in the brain by using a multi-primary stimulation system that can independently control the amount of stimulation of melanopsin cells and the other photoreceptors. We have shown that selective stimulation of melanopsin cells affects the perception of time. Furthermore, we quantified the effect on the perception of brightness from melanopsin cells.

  • Contribution of melanopsin ganglion cells to visual pathway

    Grant number:26280103  2014.4 - 2017.3

    Japan Society for the Promotion of Science  Grants-in-Aid for Scientific Research Grant-in-Aid for Scientific Research (B)  Grant-in-Aid for Scientific Research (B)

    TSUJIMURA Sei-ichi

      More details

    Grant amount:\14820000 ( Direct Cost: \11400000 、 Indirect Cost:\3420000 )

    In this study we have investigated how a stimulation of melanopsin ganglion cells contributes to visual pathway using several psychophysical and neurological techniques. We used a multi-primary stimulation system that enables us to stimulate each photoreceptor independently based on silent-substitution technique. In addition we have developed a multi-primary display system and measured contrast sensitivities as a function of spatial frequency.