LEARNER

LEARNER PUBLICATIONS

• Increasing Illumination Invariance of Learning-Based Local Features Using Photo-Realistic Simulated Environments
  Submitted at the Robotics and Autonomous Systems, Elsevier
  SSRN open-access version: download.
• A Deep Actor-Critic Reinforcement Learning Framework for Persistent Keypoint Detection under Challenging vSLAM Conditions
  Submitted at the Robotics and Automation Letters, IEEE
  Pre-print version: download
• Low-Light Adaptation for Action Recognition-Enabled Robot Navigation
  Submitted at the Robotics, MDPI
  Pre-print version: download
• MESA: A Multi-Environment Synthetic Adaptation Dataset for visual SLAM Evaluation and Feature Learning
  Presented in Proceedings of the European Conference on Mobile Robots, 2025
  Pre-print version: download
• LEARNER: A SLAM and Path Planning Middleware Package for Dynamic and Visually Challenging Environments
  Submitted at the Cyber-Systems and Robotics, IET
  Pre-print version: download