Contents Science Lab

News

November 10, 2020: Our work on Imageability estimation using visual and language features by C. Matsuhira et al. got presented at ICMR2020.

October 16, 2020: Our work on sentence imageability-aware image captioning by K. Umemura et al. got accepted by MMM2021.

April 1, 2020: Contents Science Lab was established.

Recent projects

Book page colorization based on perception

In this research, we colorize pages of books according to human perception in order to summarize and compare book contents.



Image captioning considering imageability

In this research, we incorporate psycholinguistics into image captioning in order to tailor captions to different applications.



Imageability estimation

In this research, we model the human perception of the semantic gap for use in affective computing.



Mimetic words

In this research, we analyze the phonetic characteristics and human perception of mimetic words used in Japanese to describe human gait.

Recent publications

MultiSensor-Home: Multi-modal multi-view dataset and benchmarks for action recognition in home environments. Trung Thanh Nguyen, Yasutomo Kawanishi, Vijay John, Takahiro Komamizu, Ichiro Ide. Pattern Recognition, 179(113810), pp.1-12To be published in , November 2026.

TraRA: Trajectory-level Recognition Aggregation for Video Text Spotting in Urban Surveillance. Duc Tri Tran, Trung Thanh Nguyen, Vijay John, Phi Le Nguyen, Yasutomo Kawanishi. IEEE International Conference on Advanced Visual and Signal-Based Systems, pp.1-6To be published in , September 2026.

Grouped Causal Query Adapter for Efficient Visual Adaptation in Video Captioning. Junan Chen, Trung Thanh Nguyen, Takahiro Komamizu, Ichiro Ide. Meeting on Image Recognition & Understanding, pp.1-5To be published in , August 2026.

Linear-time 3D Forest Point Cloud Segmentation with Geometry-guided Queries. Trung Thanh Nguyen, Tuan-Anh Vu, Duc Viet Le, Yasutomo Kawanishi, Takahiro Komamizu, Ichiro Ide, Teja Katternborn. Meeting on Image Recognition & Understanding, pp.1-5To be published in , August 2026.

Beyond Single-view: Cross-view Context Modeling for Multi-view Action Recognition. Trung Thanh Nguyen, Yasutomo Kawanishi, Takahiro Komamizu, Ichiro Ide. Meeting on Image Recognition & Understanding, pp.1-5To be published in , August 2026.

生成 AI 支援ロックインサーモグラフィによる 熱拡散率イメージング. 程 飛霖, シャン レイ, 鈴木 健二, 竹内 一郎, 長野 方星. Unknown Journal, June 2026.

視野が部分的に重複する多視点行動認識における視点間整合性学習法. Trung Thanh Nguyen, 川西 康友, 駒水孝裕, 井手 一郎. パターン認識・メディア理解研究会(PRMU), PRMU2025-50, March 2026.

生成AI時代に求められるデジタル人材の育成を推進. 井手 一郎. ディー・キューブ, Vol.5, p.2, March 2026.

Estimation of Cooking Proficiency based on Motion and Posture Information. Nanami Masuda, 久徳 遙矢, 道満 恵介, Takahiro Komamizu, 井手 一郎. 2nd Taiwan-Japan Multimedia Seminar (TJMM2026), A2-18, March 2026.

Image Puzzle Solving by In-context Learning of Similar-puzzle. Hayato Matsudaira, Takahiro Komamizu, 井手 一郎. 2nd Taiwan-Japan Multimedia Seminar (TJMM2026), A2-22, March 2026.