Kun-Hsiang Lin

About Me

With over seven years of experience in computer vision and AI, I am currently a Senior Computer Vision Research Engineer at Qualcomm and a Ph.D. candidate in Computer Science and Information Engineering at National Taiwan University. My academic journey began with traditional machine learning methods, including support vector machines and random forests for rainfall forecasting during my master’s studies. I later joined Academia Sinica, where I worked on deep learning–based spatial-temporal modeling and assisted with paper reviews for top-tier machine learning conferences.

Following my research at Academia Sinica, I spent over five years in industry at Authme as an AI Team Lead and CV/ML Engineer. In this role, I led the development of end-to-end AI solutions for biometric security, OCR, and fraud detection, spanning algorithm design, C++ SDK development, cloud deployment, and MLOps. My work earned top rankings in international evaluations and challenges, including the NIST FRTE benchmark and competitions held at top venues such as CVPR and ICCV.

My current research centers on leveraging vision-language large models and multimodal large language models to address a broad range of computer vision problems. I am particularly interested in instruction tuning, multimodal reasoning, model merging, and robotics, with the goal of enhancing generalization, adaptability, and interpretability in real-world AI systems.

For more information, please download my CV.

Research Interests

• Computer Vision: Image recognition, object detection, image segmentation, video understanding, anomaly detection.
• Deep Learning: Domain generalization, self-supervised learning, vision-language large models, multimodal large language models, vision-language-action models, model editing, knowledge distillation, model compression, and transfer learning.
• Biometric AI: Face recognition, face anti-spoofing, fraud detection, and deepfake detection.

News

• [Apr. 2026] Joined Qualcomm as a Senior Computer Vision Research Engineer.
• [Feb. 2026] Our paper TriDF has been accepted to the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2026.
• [Jul. 2025] Our paper on face anti-spoofing with vision-language large models has been accepted to ACM Multimedia (MM) 2025.

Education

• Ph.D. in Computer Science and Information Engineering — National Taiwan University

Experience

• Qualcomm — Senior Computer Vision Research Engineer (Apr. 2026 – present)
• Authme — AI Team Lead, Computer Vision & Machine Learning Engineer
• Academia Sinica — Research Assistant, deep learning for spatial-temporal modeling

Honors & Awards

• Best Student Paper Award — Conference on Computer Applications in Civil and Hydraulic Engineering (CCACHE), 2015

Publications

Projects

Services

Conference Reviewers

• IEEE International Conference on Multimedia and Expo (ICME) 2025
• The Thirty-Fifth International Conference on Machine Learning (ICML) 2018
• The Thirty-Second Annual Conference on Neural Information Processing Systems (NeurIPS) 2018
• The Thirty-First Annual Conference on Neural Information Processing Systems (NeurIPS, formerly NIPS) 2017
• Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD) 2017

Journal Reviewers

• IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)

Powered by Jekyll and Minimal Light theme.