Research Projects

• 

  Eyecandy Robotics RL Pipeline

  RL policies for Eyecandy Robotics — trained in Isaac Lab, validated with sim-2-sim, and run on Feetech actuators.

  Reinforcement Learning Isaac Lab Sim-2-Sim Feetech Actuators
• 
  Auto-Bot

  Half-scale autonomous robot with full kinematic simulation and embedded control.

  Robotics C++ & ROS Embedded Systems
• 
  Image Convolution on FPGA

  Hardware accelerated image processing pipeline with custom memory controllers.

  Verilog Image Processing FPGA

• 

  

• 
  MuJoCo Simulation

  Physics simulation environment for testing robotic control and RL agents.

  Python MuJoCo Robotics
• 
  NVFP4 Implementation

  NVIDIA's FP4 quantization paper, implemented for efficient edge inference.

  PyTorch Deep Learning Quantization
• 
  Hyperspectral Anomaly Detection

  Machine learning algorithms to detect anomalies in hyperspectral imaging data for remote sensing.

  Python Machine Learning Scikit-Learn