mlsanityData to 14 Oct 2026

Uncertainty-aware Grounded Action Transformation towards Sim-to-Real Transfer for Traffic Signal Control

arXiv 2307.12388 ↗First posted 23 Jul 2023cs.LG cs.AIcited by 3

Not on the map. Why?

Abstract (arXiv)

Traffic signal control (TSC) is a complex and important task that affects the daily lives of millions of people. Reinforcement Learning (RL) has shown promising results in optimizing traffic signal control, but current RL-based TSC methods are mainly trained in simulation and suffer from the performance gap between simulation and the real world. In this paper, we propose a simulation-to-real-world (sim-to-real) transfer approach called UGAT, which transfers a learned policy trained from a simulated environment to a real-world environment by dynamically transforming actions in the simulation with uncertainty to mitigate the domain gap of transition dynamics. We evaluate our method on a simulated traffic environment and show that it significantly improves the performance of the transferred RL policy in the real world.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 21 of its 39 references that are in the map. First posted on arXiv 23 Jul 2023.

Builds on

21 of 39 references.

References · 21

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,926
  3. 2018Evidential Deep Learning to Quantify Classification Uncertaintycited by 732
  4. 2018Learning Dexterous In-Hand Manipulationcited by 687
  5. 2014Deep Domain Confusion: Maximizing for Domain Invariancecited by 600
  6. 2015Variational Dropout and the Local Reparameterization Trickcited by 490
  7. 2020Sim-to-Real Transfer in Deep Reinforcement Learning for Robotics: a Surveycited by 413
  8. 2017Concrete Dropoutcited by 227
  9. 2014Robots that can adapt like animalscited by 204
  10. 2018Sim-to-Real via Sim-to-Sim: Data-efficient Robotic Grasping via Randomized-to-Canonical Adaptation Networkscited by 174
  11. 2017Using Simulation and Domain Adaptation to Improve Efficiency of Deep Robotic Graspingcited by 144
  12. 2019CoLight: Learning Network-level Cooperation for Traffic Signal Controlcited by 138
9 more
  1. 2019CityFlow: A Multi-Agent Reinforcement Learning Environment for Large Scale City Traffic Scenariocited by 115
  2. 2019Learning Phase Competition for Traffic Signal Controlcited by 87
  3. 2017Multi-Task Domain Adaptation for Deep Learning of Instance Grasping from Simulationcited by 36
  4. 2021SimGAN: Hybrid Simulator Identification for Domain Adaptation via Adversarial Reinforcement Learningcited by 27
  5. 2019Reliable deep-learning-based phase imaging with uncertainty quantificationcited by 26
  6. 2020Reinforced Grounded Action Transformation for Sim-to-Real Transfercited by 18
  7. 2020Stochastic Grounded Action Transformation for Robot Learning in Simulationcited by 16
  8. 2022LibSignal: An Open Library for Traffic Signal Controlcited by 16
  9. 2021Towards Real-World Deployment of Reinforcement Learning for Traffic Signal Controlcited by 6

Built on by

Cited by year

Citing papers per year: 2024: 2; 2025: 1.

3 papers.

  1. 2024Metacognitive AI: Framework and the Case for a Neurosymbolic Approachcited by 9
  2. 2024MalLight: Influence-Aware Coordinated Traffic Signal Control for Traffic Signal Malfunctionscited by 5
  3. 2025Joint-Local Grounded Action Transformation for Sim-to-Real Transfer in Multi-Agent Traffic Controlcited by 0