Deep Reinforcement Learning for Robotic Grasping from Octrees

reinforcement learning from randomized simulations

This project implements a simulated grasp-and-lift process in V-REP using the Barrett Hand, with an interface through a python remote API.

Novel Reinforcement Learning method for tackling goal-oriented robotics tasks with obstacles.

This is the official implementation of our work entitled "Multistream Gaze Estimation with Anatomical Eye Region Isolation by Synthetic to Real Transfer Learning" accepted in IEEE Transactions on Artificial Intelligence

DROPO: Sim-to-Real Transfer with Offline Domain Randomization

[WACV 2024] AnyStar: Domain randomized universal star-convex 3D instance segmentation

The repository is intended as a support tool for the report of the project "Sim to Real transfer of Reinforcement Learning Policies in Robotics" and it contains examples of some well-known algorithms and methods in the fields of Reinforcement Learning and Sim-to-Real transfer. The implementation is not thought to be efficient, thus we suggest yo…

Enabling Faster Training of Robust Reinforcement Learning Policies for Soft Robots

Automatic Domain Randomization (ADR) proposed in "Solving Rubik's Cube with a Robot Hand"

A synthetic data generator for retail object detection, which is built with Blender and Python.

Development of an image-based autonomous driving system for an e-FSAE.

A simple framework to change visual and physical parameters of various simulations without reloading the simulation and through a standardized interface.

Using Blender and Python to generate synthetic data for training an 8-class YOLOv5 detector in a highway driving scenario (Sedan, SUV, Van, Truck, Bus, Warning Sign, Speed Limit Sign, Speed Cam Sign).

Augmentation of event-based machine learning dataset for image segmentation using noise injection

Sim2Real for joint robotic locomotion and manipulation with RCAN

Generates robot models with noise added to URDF parameters for domain randomization

Code for the paper "Generative Adversarial Reinforcement Learning"