Scene Segmentation using UNet-based architecture to teach the robot to understand what's in an image by dividing it into different sections and identifying what's in each section.
Computer Vision / Deep LearningÂ
An autonomous robot designed to detect human poses, aimed at assisting the elderly population, is deployed on Jetson Orin to ensure swift inference for real-time performance.
Computer Vision / Edge Machine Learning
Perception-aware path planning using MPPI (Model Predictive Path Integral) involves finding the best route for a robot or vehicle while considering what it "sees" in its environment.
Computer Vision / Reinforcement Learning