polrizzo

LSTA-3N

Two well-known solutions have been simultaneously implemented for the first time, in order to mitigate two of the biggest challenges which come with egocentric videos when performing action recognition. TA3N has been enriched with LSTA architecture.

CityFlow

A Multi-Agent Reinforcement Learning Environment for Large Scale City Traffic Scenario

colight

CoLight: Learning Network-level Cooperation for Traffic Signal Control

marl_cityflow

NetworkDynamics

Repository for the Network Dynamics and Learning Course @ PoliTo

Recommendation_System_KRED

Implementation of Multi-Head Attention and Enhancement of Entity Embeddings on KRED: Knowledge-Aware Document Representation for News Recommendations.