Silvester Hsu

I am a developer dedicated to efficient algorithms and GPU programming, with a focus on research and application in computer vision, deep learning, and high-performance computing. I have extensive experience in parallel computing, image processing, and inference engine optimization. I specialize in performance optimization using CUDA and have a deep understanding of GPU memory management and acceleration of deep learning frameworks.

In my GitHub repository, you can find open-source projects across multiple domains, ranging from GPU memory management tools and image distortion correction algorithms to efficient model inference engine optimizations. I have contributed code to several deep learning projects and successfully improved inference speed and computational efficiency.

Tech Stack

• Programming Languages: C++, Python, Bash
• GPU Programming: CUDA, TensorRT, cuDNN
• Deep Learning Frameworks: TensorFlow, PyTorch, ONNX
• Image Processing: OpenCV, Computer Vision (CV)
• Databases: SQLite
• Operating Systems: Linux (Ubuntu), Windows
• Tools & Environments: Git, Docker, Jenkins, CMake
• Performance Optimization: SIMD, Multithreading, High-Performance Computing (HPC)

I have a strong interest in algorithm optimization, computational graph design, and GPU acceleration, particularly in applications such as autonomous driving and image recognition. I look forward to connecting and collaborating with like-minded developers to push the boundaries of cutting-edge technology.