Skip to content
View IonThief's full-sized avatar

Block or report IonThief

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
IonThief/README.md

Computer Vision Engineer

Skills

Python C++ CUDA PyTorch OpenCV Docker


Open Source Contributions


Completed Projects

  • [TBD]

Planned Projects

Systems & C Programming

  • Custom Memory Allocator: Implementation of malloc and free to manage heap memory at the OS level.
  • Multithreaded HTTP Server: A concurrent web server handling socket connections and network protocols.
  • Minimal Linux Shell: A command-line interpreter demonstrating process creation and signal handling.

Graphics & OpenGL

  • Custom Software Rasterizer: A 3D rendering pipeline written from scratch without GPU acceleration.
  • Voxel Engine: A block-based rendering engine focusing on chunk management and greedy meshing.
  • Real-Time Fluid Simulation: An Eulerian fluid solver implemented using compute shaders.
  • Particle System: A GPU-accelerated engine simulating physical interactions of millions of particles.

Computer Vision & Deep Learning

  • Camera Calibrator: A mathematical solver to compute intrinsic parameters and remove lens distortion.
  • C++ Lane Tracker: Finding and tracking road lines in real-time video.
  • Visual Odometry System: Trajectory estimation using sequential camera frames.
  • Hand Gesture Controller: Moving the computer mouse using hand shapes and movements.
  • Face Anonymizer: A fast inference pipeline for real-time privacy blurring in videos.
  • ViT Image Classifier: Using attention mechanisms to see what parts of a picture the AI focuses on.
  • Transformer Object Tracker: Bounding box regression using attention to track moving objects over time.
  • Cross-Modal Attention Engine: A tool that links text descriptions to the correct parts of an image.

Robotics & Sensor Fusion

  • ROS 2 Vision Node: A distributed component for streaming and processing live sensor data.
  • Event Camera Tracker: High-speed motion estimation using asynchronous event-based sensors.
  • LiDAR-Camera Fusion: Cross-calibrating 3D point clouds with 2D RGB data for dense depth estimation.
  • LLM-to-ROS Commander: Typing commands in plain English to make a robot move.
  • RL Navigation Agent: A reinforcement learning policy for autonomous obstacle avoidance.

Hardware Optimization & Deployment

  • CUDA Point Cloud Filter: Cleaning up messy 3D sensor scans instantly using the graphics card.
  • TensorRT Optimizer: Squishing heavy AI models so they run fast on tiny edge devices.
  • Dockerized Vision App: Wrapping up a complete vision tool so anyone can run it with one command.
  • Factory Defect Detector: A lightweight model that spots scratches on products quickly for an assembly line.
  • NLP Log Analyzer: A script that automatically reads machine error text to figure out why the robot broke.

Pinned Loading

  1. chrischoy/DeepGlobalRegistration chrischoy/DeepGlobalRegistration Public

    [CVPR 2020 Oral] A differentiable framework for 3D registration

    Python 542 88

  2. spikingjelly spikingjelly Public

    Forked from fangwei123456/spikingjelly

    SpikingJelly is an open-source deep learning framework for Spiking Neural Network (SNN) based on PyTorch.

    Python