Hi, I'm He-Yen (Louis) Hsieh, a Ph.D. candidate in Computer Science at Harvard University, advised by Prof. H. T. Kung.
I enjoy building machine learning algorithms that make AI models smaller, faster, and easier to adapt for real-world deployment. My work spans efficient AI, generative AI, large language models, and computer vision, with recent projects in model adaptation, post-training quantization, knowledge distillation, efficient inference, gaze estimation, and personalization.
What I enjoy most is turning research ideas into working AI systems: designing new algorithms, building prototypes, and making new ideas practical.
Before and during my Ph.D., I have been fortunate to work with researchers at Intel, Reality Labs Research at Meta, Adobe Research, and Academia Sinica. My work has been published at CVPR, AAAI, ECCV, WACV, ICIP, and ICASSP.
I'm always happy to connect with people interested in efficient AI, model compression, generative models, or research ideas for making AI smaller, faster, easier to adapt, and still accurate.
Visualizing Ideas
I enjoy using visual diagrams, animations, and interactive demos to distill technical ideas so they are easier to understand. For example, during my time as a teaching assistant, I built a step-by-step visualization of Dijkstra’s algorithm to explain graph search. I also like making research more accessible through demos, including GazeGen and Gen4Gen.
Step-by-step visualization of Dijkstra's algorithm.
Research Roadmap
My research centers on making AI practical under limited data, compute, and memory. I started from computer vision problems such as detection, video understanding, and anomaly detection, then expanded into gaze estimation, generative AI, and efficient learning systems. More recently, my work focuses on efficient adaptation and compression for vision and language models.
Efficient Model Adaptation
Alfa — Efficient few-shot model adaptation
2026Adapts gaze models to new users and domains with only 5 samples by reusing pre-trained knowledge, making the model 5× smaller while keeping or improving accuracy. The same low-rank adaptation idea also extends to diffusion-based language models for reasoning.
DFT Gaze — Few-shot gaze personalization
2025Compresses gaze estimation to a tiny 281K-parameter model, keeping 90% of the precision of a 10× larger model while supporting real-time personalization on edge devices.
Efficient AI Deployment
Calibration-free PTQ for LLMs
OngoingMakes 3-/4-bit LLM deployment easier when calibration data is unavailable or hard to collect.
DFT Gaze — Compact vision model design
2025Compresses gaze estimation to a tiny 281K-parameter model, keeping 90% of the precision of a 10× larger model while supporting real-time personalization on edge devices.
EDIT — Faster diffusion language model inference
2025Speeds up diffusion-based language model reasoning by stopping inference once generation has stabilized.
- Diffusion-based language models use more inference steps than needed.
- Reasoning tasks need faster generation under test-time compute budgets.
Generative AI
Personalized text-to-image generation
2023Few-shot personalized generation from only a few user-provided examples.
- Personalized generation is needed from only a few examples.
- Multiple user-provided concepts must appear in one generated image.
Gen4Gen — Generative AI for training data synthesis
2025Builds a generative pipeline to create diverse synthetic training data for stronger downstream vision models.
- Generative models can be used to build a data synthesis pipeline.
- Real training data is limited, expensive, or hard to collect.
GazeGen — Gaze-guided generation
2024Turns gaze into a natural control signal for interactive image and video generation.
- User intent should guide generation without heavy manual interaction.
- Gaze can serve as a lightweight control signal for interactive AI systems.
Computer Vision
S3R — Video anomaly detection
2022Detects abnormal events in videos with weaker supervision, reducing the need for heavy manual annotation.
- Video understanding needs to scale without dense frame-level labels.
- Rare events must be detected from compact normal-pattern representations.
Adaptive Image Transformer — One-shot object detection
2021Recognizes new object categories from only one example, making detection more flexible.
- New object categories must be detected from only one example.
- Full training data is unavailable for every new class.
Temporal action detection
2020–2023Finds actions in long videos from only one example, using zoom-in attention to refine coarse-to-fine action locations.
- Long videos need action localization with limited examples.
- Coarse video cues must be refined into precise temporal action boundaries.
How Can AI Work with Less?
Choose a constraint and see how my projects help an existing AI model work with less.