This is a research simulator for swarm robotics. It uses client-server architecture with asynchronous socket communication. It can be used to run algorithms like flocking and grain separation.
This is a simulator for 3D robot evolution, using a parallel hill climber algorithm with random mutation. Robots are generated using a branching algorithm.
This is a CDCGAN and Double Deep Q-Network trained to generate and play original levels of Super Mario Bros 2. The GAN was trained on a subset of the VGLC and the RL agent was trained on GAN output.
This is a web service where users can upload an image of a garbage item and see whether or not the item is recyclable. It uses AWS and consists of a client, an EC2 server, an S3 bucket, and a Lambda function that triggers a Tensorflow model.
