Accepted to 2022 IEEE GLOBECOM Workshops

Our paper has been accepted to the Workshop on Edge Learning over 5G Mobile Networks and Beyond in IEEE GLOBECOM 2022). This work is collaborative research with Prof. Nishio at Tokyo Institute of Technology. This paper proposes a neural architecture search method for split computing in which neural networks are split and located at edge devices and servers for inference.

  • Shoma Shimizu, Takayuki Nishio, Shota Saito, Yoichi Hirose, Chen Yen-Hsiu, and Shinichi Shirakawa: Neural Architecture Search for Improving Latency-Accuracy Trade-off in Split Computing, 2022 IEEE Globecom Workshops, Edge Learning over 5G Mobile Networks and Beyond, Rio de Janeiro, Brazil, December 4-8, 2022. [arXiv]