About Me

Hi, I’m Shih-Yao (Mike) Lin, currently a Senior Applied Scientist at Amazon Prime Video. I work on post-training foundation multimodal LLMs and video understanding, with a focus on areas like VLM-based video action prediction, captioning, and audio-based profanity language detection. I’m particularly interested in building practical AI systems that help better understand and moderate video content.

Previously, I was part of the Amazon AGI team, contributing to the pretraining of large language and vision-language models. Before that, I worked at Sony (2021–2022) as a Senior Research Scientist, where I developed 3D character motion for real-time game engines. From 2018 to 2021, I was a Senior Researcher at Tencent America, focusing on human pose estimation and motion analysis. Earlier in my career, I held a postdoctoral role at Bosch (2017–2018), working on sensor-based 3D hand gesture recognition.

My work bridges multimodal learning and computer vision, with a focus on building scalable, impactful AI systems through interdisciplinary collaboration.

Photo: Standing before Mount Rainier — a quiet moment that mirrors my work in AI: navigating complexity, appreciating scale, and always aiming higher.