[WP-E] Poster Session 2

[WP-E-10] Fundamental Investigation of Performance Degradation of Multimode Optical Fiber Imaging with Variational Autoencoder against Fiber Bending

○Taiki Otani¹, Masatoshi Bunsen², Hana Fowler¹ (1. Graduate School of Engineering. Fukuoka University (Japan), 2. Fukuoka University (Japan))

Keywords:

Emerging non-telecom applications including sensing and imaging,Other novel fibers,fiber devices and applications,Artificial intelligence and machine learning for optical transmission systems and subsystems

In multimode optical fiber imaging with variational autoencoder (VAE), deformation of the fiber after VAE training can significantly degrade image recovery performance. As a first step to solve this problem, we investigate the extent and process of this degradation for intentional fiber bending by optical experiments.