ROS 2 Offline Voice Assistant
Kilo Bot listens continuously with Vosk, runs a local Llama model through Ollama and speaks with Piper. Voice activity detection lets it stop speaking when interrupted.
View Project →Designing intelligent automation and autonomous systems that redefine the intersection of AI and physics.
Hello! I am Hmue Set San, a professional Robotics & AI Developer with a rich academic foundation from three premier technical institutions.
My journey began with deep-diving into computer science, spending 1.5 years at Computer University (Sittwe), 1.5 years at Computer University (Monywa), and 1 year at Computer University (Pathein). This academic rotation provided me with a unique perspective on diverse technical ecosystems and research methodologies.
Currently, I architect intelligent robotic solutions and machine learning models, merging complex algorithms with physical automation to solve real-world challenges.
Here are some technologies I work with:
Kilo Bot listens continuously with Vosk, runs a local Llama model through Ollama and speaks with Piper. Voice activity detection lets it stop speaking when interrupted.
View Project →Real-time webcam emotion inference using MediaPipe face detection and an EfficientNet-V2-S classifier trained with PyTorch. Includes a training notebook and evaluation results.
View Project →A ROS 2 four-wheel car robot model with separate robot-description and Gazebo simulation packages.
View Project →Explore images and live camera feeds with local vision-language models through a browser UI, with inference options for Transformers, ONNX Runtime and llama.cpp.
View Project →Run YOLOv8 detection on images, videos or webcam feeds using ONNX or PyTorch models, with CPU and CUDA setup options.
View Project →Feel free to reach out for collaborations or just a friendly hello!