[TuG3] Algorithms for Photonic Computing

2:00 PM - 2:15 PM JST(5:00 AM - 5:15 AM UTC)

[TuG3-1] Simple Machine Learning Using Multi-Port Optical Directional Couplers: Comparing Coherent and Non-Coherent Approaches

○Yushan Hu¹, Keigo Takabayashi¹, Ryota Nakayama¹, Satoshi Sunada¹, Takeo Maruyama¹ (1. Kanazawa University (Japan))

Keywords:

Deep learning for photonic device and applications,Photonics in neuromorphic computing and machine learning devices

Optical neural networks are emerging as structures with potential for artificial intelligence. We describe a multi-port directional coupler architecture and demonstrate its application for classification tasks, achieving different accuracies across different experimental conditions.