Image Instance Segmentation Tensorflow Python Evaluated and designed state-of-the-art algorithms for image instance segmentation. Used various methods such as image augmentation and RGB-D to achieve high accuracy. Familiarized with machine learning research methodologies such as keeping detailed research logs. This was used as part of a robotics picking pipeline...
Robotic Bin Picking Tensorflow Python Bash Worked on deep learning algorithms for generating and evaluating grasping points for picking objects in occluded scenes. Designed data generation strategies with simulation software. Designed automated tools with Python and Bash scripts to evaluate the accuracy and success rate of the grasping point detection...
Realistic Robotic Simulation Python Lua Developed Python and Lua scripts to realistically simulate robot grasping for training data generation with V-REP and the PyBullet physics engine.