Presentation Information

[9a-N106-6]Development and Application of Nanomaterials Based In-sensor Reservoir Devices

Kouki Kimizuka1, Saman Azhari1,2, Shshi Tokuno1, Ahmet Karacali1, Yuki Usami1,2, Shuhei Ikemoto1,2, Hakaru Tamukoh1,2, 〇Hirofumi Tanaka1,2 (1.LSS, 2.Neumorph Center, Kyushu Inst. Tech.)

Keywords:

Reservoir computing devices,robot application,hardware development

Reservoir computing (RC) is a computational algorithm that enables high-speed information processing and demonstrates excellent performance, particularly in time-series data processing and pattern recognition. Physical RC computing devices have seen explosive growth in recent years due to high expectations for their use in AI hardware. Here, we used carbon nanotube sponge sensors to perform object recognition and achieved a recognition rate of over 80%. We also attempted real-time recognition, which we will introduce at the presentation.