Keynote Session
9:10-10:45 at Conference center
9:10-9:45 (35 min) Keynote Speech A
Prof. Seiichi Uchida, Kyushu University, Japan
- https://human.ait.kyushu-u.ac.jp/index-e.html
- "Typographic design meets AI"
- Typography -- the design of letters -- affects not only readability but also emotion and impression. Fonts can signal trust, playfulness, or elegance, yet such effects are often judged intuitively. This talk shows how AI enables systematic analysis of typography, linking visual features with perception and supporting applications such as font recommendation. I will highlight insights from my research into the relationship between form and meaning.
9:45-10:20 (35 min) Keynote Speech B
Prof. Makoto Tsubokawa, Waseda University, Japan
- https://www.waseda.jp/fsci/gips/other-en/2015/09/08/2161/
- "Exploring Lateral Input/Output Techniques for Optical Waveguides"
- In optical waveguides, the optical input/output interface is usually the end faces. However, converting this interface to a top or side surface is not straightforward, and there are few examples of actual research on the topic. This technique is attractive, however, and has the potential to be used in a wide range of applications, from relatively large thin-film solar cells to fine photonic integrated circuits. The author has previously researched lateral input/output methods in millimeter- to nanoscale optical fibers and waveguides. Due to their flexibility and long-length capability, large millimeter-scale optical fibers are used in light concentrators, flexible sensors, and photonic textile displays. In addition, applications at the submicron scale are primarily used for interconnections between microdevices on photonic integrated circuits and for coupling with optical fibers, surface-illuminated detectors, and surface-emitting lasers. The advent of nanoscale processing technology has enabled the creation of complex structures called metasurface, leading to the emergence of increasingly diverse devices. In this presentation, I will present some examples from our previous research.
10:20-10:35 Break
10:35-11:10 (35 min) Keynote Speech C
Prof. Song Chen, University of Science and Technology of China, China
- http://staff.ustc.edu.cn/~songch/
- "DRAM-based Processing-in-Memory Systems: Progress, Challenges, and Opportunities"
- Along with the development of 3D integration technology, the DRAM-based Processing-in-Memory (DRAM-PIM) systems have become a promising solution for accelerating deep neural networks, including CNNs and transformer networks. In this keynote speech, we will provide a brief review of the progress in DRAM-PIM systems and discuss the opportunities and challenges for accelerating LLM inference with DRAM-PIM, from the perspectives of architecture, compiler, and implementation.
11:10-11:45 (35 min) Keynote Speech D
Prof. Daniel T. Simon, Linköping University, Sweden
- https://liu.se/en/research/laboratory-of-organic-electronics/organic-bioelectronics
- "Bridging the gap: Organic Bioelectronics at Linkoping University"
- The Organic Bioelectronics group merges electronics and biology, pioneering iontronic and in situ bioelectronics to create minimally invasive, high-precision devices that integrate with living systems. We advance neuromodulation, drug delivery, and biosensing. This talk presents our work on self-organizing in vivo electrodes and on-demand chemical delivery to address complex disorders, from nervous system diseases to cancer.
