About Me. I am a tenure-track Assistant Professor of Computer Science at Boise State University. My research focuses on computer vision, domain adaptation, adaptive foundation models, training efficiency, and their real-world applications. I bring extensive industry experience to my academic role. Prior to joining Boise State, I worked as a Machine Learning Engineer at Bastian Solutions, a Toyota Automated Logistics company, where I developed machine learning solutions for complex problems in computer vision and robotics. I also completed an internship at Siemens, focusing on AI for healthcare. I earned my Ph.D. in Computer Science from the University of Kentucky in March 2023, under the supervision of Dr. Nathan Jacobs. In my free time, I enjoy outdoor activities across Idaho and the broader Pacific Northwest.

Lab Openings. I am actively seeking for new members at all levels to join my Lab! We encourage a supportive culture where you can do impactful research, learn new skills, and achieve your personal goals.

Research Interests: My research aims on adaptive and efficient perception and decision-making for real-world autonomy. I am interested in a wide range of research fields from computer vision, domain adaptation, adaptive foundation models, training efficiency, robotics, remote sensing, and AI for Interdisciplinary Sciences. Currently I am actively working on:

1. Domain Adaptive Computer Vision.
   The appearance of images can vary significantly due to factors such as lighting conditions, time of day, weather, viewpoint, and seasonal changes. These variations must be carefully considered and modeled when designing robust vision systems. I often refer to this challenge as a multi-domain adaptation problem. It presents a wide range of research opportunities, including unsupervised, semi-supervised, and few-shot learning approaches, as well as methods guided by large language models (LLMs) and vision-language models (VLMs), to train vision systems that are both robust and domain-adaptive.
   Keywords: Domain Adaptation, Computer Vision, LLMs, VLMs
2. Computer Vision in 3D and Robotics.
   Understanding the 3D structure of the world is fundamental for enabling intelligent robotic systems. Core 3D computer vision tasks such as depth estimation, pose regression, reconstruction, and point cloud understanding play a critical role in how robots perceive and interact with their environment. Advancing research in this area creates opportunities to explore self-supervised learning, multi-view and multi-modal fusion, uncertainty estimation, real-time 3D perception, and ultimately, achieving real-world autonomy.
   Keywords: Robotics, 3D Vision, Point Cloud, Self-Supervised Learning
3. Efficient and Secured Machine Learning and Edge AI.
   Designing efficient vision systems for Edge AI requires models that are compact, fast, and secured. Techniques such as quantization and pruning help reduce model size and computational cost, making real-time inference feasible on edge devices. Federated learning further enables distributed training without sharing raw data, enhancing privacy and scalability. Together, these approaches open up research opportunities in building lightweight, adaptive, and secured AI models that operate effectively across diverse and decentralized environments.
   Keywords: Federated Learning, Efficient AI, AI Safety, Edge AI

Collaboration with Me. I am open to research collaborations, invited talks, and other opportunities from both within and outside my institution. I believe that impactful research often grows from long-term, meaningful collaborations. If you're interested in working together or starting a conversation, feel free to drop me an email. Let’s connect!

---

Recent News

---