Hiroki Kobayashi
Computer Vision Researcher & AI Engineer

About

Hiroki Kobayashi

I am an AI engineer at Araya Inc. with a background in computer vision research, specializing in industrial anomaly detection, representation learning, and generative modeling.

My work focuses on practical and robust machine learning methods for visual inspection, especially in manufacturing settings where defective samples are scarce and strong generalization matters more than benchmark cosmetics.

I am particularly interested in pseudo-defect generation, domain-specific pre-training, feature modeling for anomaly detection, and bridging the gap between academic research and deployable AI systems.

Beyond research, I actively develop deep learning implementations in both Python and C++, including large-scale open-source work built on PyTorch and LibTorch.

↓ Download CV (PDF)
Current AI Engineer, Araya Inc.
Since April 2026
  • The Institute of Electronics, Information and Communication Engineers (IEICE)
  • IEEE
Apr. 2026 – Present AI Engineer, Araya Inc.
Apr. 2023 – Mar. 2026 PhD in Engineering, Chukyo University — research on industrial anomaly detection, pseudo-defect generation, and domain-specific pre-training.
Apr. 2021 – Mar. 2023 MSc in Engineering, Chukyo University — research on unsupervised defect detection based on normal image generation.
Apr. 2017 – Mar. 2021 BSc in Engineering, Chukyo University — research on scratch detection using realistic image generation from before/after defect pairs.
Anomaly Detection Computer Vision Visual Inspection Representation Learning Generative Modeling PyTorch LibTorch Industrial AI