日本語ページはこちら.

News & Topics

Our paper for Sensors was accepted. (2023/8/30)
paper bibtex
We gave a poster presentation at CVPR2022.(2022/6/24)
CVPR2022 Poster
We received the MIRU Nagao Award (first prize) at MIRU2021. MIRU2021 Award

Name

Michitaka Yoshida

Organization

Research Fellow of the Japan Society for the Promotion of Science (PD), Shizuoka University

Biographies

Michitaka Yoshida received his B.E. and M.E. degrees from Kyushu University, Fukuoka, Japan, in 2017 and 2019, and Ph.D. degree from Graduate school of Information Science and Technology, Osaka University, Suita, Japan in 2023. He currently works at Shizuoka University as a research fellow (PD) of the Japan Society for the Promotion of Science. His research interests are computational photography and computer vision.

Contact

E-mail : yoshida.michitaka[at]shizuoka.ac.jp

Publications

Google Scholar

ORCID

Journals

[Peer-reviewed]
Ryoya Mizuno, Keita Takahashi, Michitaka Yoshida, Chihiro Tsutake, Toshiaki Fujii, and Hajime Nagahara,
“Compressive Acquisition of Light Field Video Using Aperture-Exposure-Coded Camera”,
ITE Transactions on Media Technology and Applications 2024, Vol.12, Issue 1, pp.22-35. paper bibtex
[Peer-reviewed]
Yoshida Michitaka, Akihiko Torii, Masatoshi Okutomi, Rin-ichiro Taniguchi, Hajime Nagahara, and Yasushi Yagi,
“Deep Sensing for Compressive Video Acquisition”,
Sensors 2023, 23, no. 17: 7535. paper bibtex
[Peer-reviewed]
Sudhakar Kumawat, Tadashi Okawara, Michitaka Yoshida, Hajime Nagahara, Yasushi Yagi,
“Action Recognition From a Single Coded Image”, IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, no. 4, pp. 4109-4121, 1 April 2023, doi: 10.1109/TPAMI.2022.3196350.
paper bibtex
[Peer-reviewed]
Michitaka Yoshida, Toshiki Sonoda, Hajime Nagahara, Kenta Endo, Yukinobu Sugiyama, Rin-ichiro Taniguchi,
“High-Speed Imaging Using CMOS Image Sensor With Quasi Pixel-Wise Exposure”,
IEEE Transactions on Computational Imaging, Vol.6, pp.463-476, 2019
paper bibtex

International Conferences

[Peer-reviewed, Poster] Michitaka Yoshida, Daisuke Hayashi, Lioe De Xing, Keita Yasutomi, Shoji Kawahito, Keiichiro Kagawa, Hajime Nagahara, “Single-shot efficient depth imaging based on time-compressive CMOS image sensor”,
Proceedings of the IEEE Sensors, Kobe, Japan, 2024.10
[Non-peer-reviewed, Poster]
Michitaka Yoshida, Daisuke Hayashi, Lioe De Xing, Keita Yasutomi, Shoji Kawahito, Keiichiro Kagawa, Hajime Nagahara, “Single-shot efficient depth imaging based on time-compressive CMOS image sensor”,
Proceedings of the International Conference on Computational Photography (ICCP2024), Lausanne, Switzerland, 2024.07
[Non-peer-reviewed, Oral]
Michitaka Yoshida, Daisuke Hayashi, Lioe De Xing, Keita Yasutomi, Shoji Kawahito, Keiichiro Kagawa, Hajime Nagahara, “Single-shot efficient transient imaging with optimal coded shutter based on time-compressive CMOS image sensor”,
Computational Optical Sensing and Imaging (COSI), Toulouse, France, 2024.07
[Peer-reviewed, Poster]
Ryoya Mizuno, Keita Takahashi, Michitaka Yoshida, Chihiro Tsutake, Toshiaki Fujii, Hajime Nagahara, “Acquiring a Dynamic Light Field Through a Single-Shot Coded Image”,
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), New Orleans, U.S.A, 2022.06
[Peer-reviewed, Oral]
Tadashi Okawara, Michitaka Yoshida, Hajime Nagahara, Yasushi Yagi,
“Action Recognition from a Single Coded Image”,
Proceedings of the International Conference on Computational Photography (ICCP2020), Saint Louis, U.S.A, 2020.04
[Non-peer-reviewed, Poster]
Michitaka Yoshida, Akihiko Torii, Masatoshi Okutomi, Kenta Endo, Yukinobu Sugiyama, Rin-ichiro Taniguchi, Hajime Nagahara,
“Joint optimization for compressive color video sensing and reconstruction under hardware constraints”,
Proceedings of the International Conference on Computational Photography (ICCP2020), Saint Louis, U.S.A, 2020.04
[Non-peer-reviewed, Demonstration]
Michitaka Yoshida, Akihiko Torii, Masatoshi Okutomi, Kenta Endo, Yukinobu Sugiyama, Rin-ichiro Taniguchi, Hajime Nagahara,
“Joint optimization for compressive video sensing and reconstruction under hardware constraints”,
Proceedings of the International Conference on Computational Photography (ICCP2019), Tokyo, Japan, 2019.05
[Non-peer-reviewed, Poster]
Michitaka Yoshida, Akihiko Torii, Masatoshi Okutomi, Kenta Endo, Yukinobu Sugiyama, Rin-ichiro Taniguchi, Hajime Nagahara,
“Joint optimization for compressive video sensing and reconstruction under hardware constraints”,
Proceedings of the 4th International Workshop on Image Sensors and Imaging Systems (IWISS2018), Tokyo, Japan, 2018.11
[Peer-reviewed, Poster]
Yoshida Michitaka, Torii Akihiko, Okutomi Masatoshi, Endo Kenta, Sugiyama Yukinobu, Taniguchi Rin-ichiro, Nagahara Hajime,
“Joint optimization for compressive video sensing and reconstruction under hardware constraints”,
Proceedings of the European Conference on Computer Vision (ECCV2018), Munich, Germany, pp.649-663, 2018.09
paper bibtex

Domestic Conferences

[Non-peer-reviewed, Poster]
吉田道隆, 林大輔, リュウドゥシン, 安富啓太, 川人祥二, 香川景一郎, 長原一, “超高時間分解コンピュテーショナルCMOSイメージセンサを用いたシングルショットTime of Flight Imaging”,
第27回画像の認識・理解シンポジウム(MIRU2024), 熊本, 2024年8月
[Non-peer-reviewed, Demonstration]
林大輔, 吉田道隆, De Xing Lioe, 安富啓太, 川人祥二, 長原一, 香川景一郎, “電荷領域符号化露光ToF CMOSイメージセンサと深層学習モデルによるトランジェントイメージングと距離推定システム”,
第27回画像の認識・理解シンポジウム(MIRU2024), 熊本, 2024年8月
[Non-peer-reviewed, Oral] 吉田道隆, ファムゴックアン, リュウドゥシン, 安富啓太, 川人祥二, 香川景一郎, 長原一, “超高時間分解コンピュテーショナルCMOSイメージセンサを用いた深層学習による疑似dToF”, 映情学技報, vol. 47, no. 27, IST2023-37, pp. 9-12, 2023年9月.
[Non-peer-reviewed, Oral] 吉田道隆, 長原一, “圧縮センシングによる動画像の高速撮像に関する研究”, 情報処理学会コンピュータビジョンとイメージメディア研究会(2023-CVIM-234), 名古屋, 2023年5月
[Non-peer-reviewed, Oral] 加藤睦代, 岩崎昌子, 長原一, 吉田道隆, 末原大幹, 山田悟, 中島悠太, 武村紀子, 中野貴志, “機械学習を用いたスパースサンプリングによるデータ処理技術の基礎開発(II)”, 日本物理学会2022年秋季大会, 岡山 2022年9月
[Peer-reviewed, Oral] 水野良哉, 高橋桂太, 坂井康平, 都竹千尋, 藤井俊彰 , 吉田道隆, 長原一, “動的光線空間のシングルショット撮影”, 第24回画像の認識・理解シンポジウム(MIRU2021), 名古屋, 2021年7月
[Non-peer-reviewed, Oral] 宮崎袈, 渡邉晃平, 吉田道隆, 安富啓太, 川人祥二, 長原一, 香川景一郎, “XY画素アドレスを用いた圧縮ビデオイメージセンサ”, 映情学技報, vol. 45, no. 17, IST2021-34, pp. 17-20, 2021年6月.
[Non-peer-reviewed, Oral]
大河原忠,吉田道隆, 長原一, 八木康史,
“符号化露光画像を用いた人物の行動認識”,
情報処理学会コンピュータビジョンとイメージメディア研究会(2019-CVIM-220), 奈良, 2020年1月
[Non-peer-reviewed, Demonstration]
吉田道隆, 鳥居秋彦, 奥富正敏, 遠藤健太, 杉山行信, 谷口倫一郎, 長原一,
“DNNにより最適化されたピクセルコーディングCMOSイメージセンサによるハイスピード撮像”,
第22回画像の認識・理解シンポジウム(MIRU2019), 大阪, 2019年7月
[Non-peer-reviewed, Poster]
大河原忠, 吉田道隆, 長原一, 八木康史,
“符号化露光画像を用いた人物の行動認識”,
第22回画像の認識・理解シンポジウム(MIRU2019), 大阪, 2019年7月
[Non-peer-reviewed, Oral]
大河原忠, 吉田道隆, 長原一, 八木　康史,
“符号化露光画像を用いた人物の行動認識”,
情報処理学会コンピュータビジョンとイメージメディア研究会(2018-CVIM-215), 京都, 2019年1月
[Non-peer-reviewed, Poster]
吉田道隆, 鳥居秋彦, 奥富正敏, 遠藤健太, 杉山行信, 谷口倫一郎, 長原一,
“ハードウェアの制約を考慮した圧縮ビデオセンシングにおける圧縮と再構成の同時最適化”,
第21回画像の認識・理解シンポジウム(MIRU2018), 札幌, 2018年8月
[Non-peer-reviewed, Oral]
吉田道隆, 長原一, 鳥居秋彦, 奥富正敏, 谷口倫一郎,
“Deep Learningによる圧縮ビデオセンシングの再構成”,
情報処理学会コンピュータビジョンとイメージメディア研究会(2017-CVIM-208), 東京, 2017年9月
[Non-peer-reviewed, Poster]
吉田道隆, 長原一, 鳥居秋彦, 奥富正敏, 谷口倫一郎,
“圧縮センシング動画のデコーディング手法の検討”,
第20回画像の認識・理解シンポジウム(MIRU2017), 広島, 2017年8月

Others

[Invited talk]
“Deep compressive sensing with coded image sensor”, Michitaka Yoshida, Daisuke Hayashi, Lioe De Xing, Keita Yasutomi, Shoji Kawahito, Keiichiro Kagawa, Hajime Nagahara, 6th International Workshop on Image Sensors and Imaging Systems (IWISS2024), Tokyo, Japan, 2024.11
[Invited talk]
Forum on Information Technology (FIT2019) :
Yoshida Michitaka, Torii Akihiko, Okutomi Masatoshi, Endo Kenta, Sugiyama Yukinobu, Taniguchi Rin-ichiro, Nagahara Hajime,
“Joint optimization for compressive video sensing and reconstruction under hardware constraints”,
FIT 2019, Okayama, Japan, 2019.09
[Patent]
特許第7272625号，「動画像処理方法及び動画像処理装置」,長原一, 大河原忠, 吉田道隆 J-GLOBAL, J-PlatPat

Research

Compressive video sensing

Compressive video sensing is a task to encode multiple sub-frames into a single frame with controlled sensor exposures and to reconstruct the sub-frames from the single compressed frame. It is known that spatially and temporally random exposures give the most balanced compression in terms of signal recovery. However, fabricating the sensors that can arbitrary expose on each individual pixel is infeasible. It is, therefore, necessary to design the exposure pattern while taking into account the constraints induced by the hardware. In this work, we propose a method for jointly optimizing, under the hardware constraints, the exposure patterns of compressive sensing and the reconstruction framework.