Learning Based
Learning-based approaches are revolutionizing various fields by enabling systems to learn complex behaviors and adapt to dynamic environments, primarily aiming to improve efficiency, robustness, and safety. Current research focuses on applying deep reinforcement learning, diffusion models, and Koopman operators to control robots, optimize planning algorithms (like those for pathfinding and task sequencing), and improve the accuracy and efficiency of simulations. These advancements have significant implications for robotics, autonomous systems, and other domains requiring adaptable and intelligent control, offering solutions to challenges in areas such as safe navigation, precise manipulation, and efficient resource allocation.
Papers
Learning Koopman Dynamics for Safe Legged Locomotion with Reinforcement Learning-based Controller
Jeonghwan Kim, Yunhai Han, Harish Ravichandar, Sehoon Ha
Neural Control Variates with Automatic Integration
Zilu Li, Guandao Yang, Qingqing Zhao, Xi Deng, Leonidas Guibas, Bharath Hariharan, Gordon Wetzstein
Low-Cost Tree Crown Dieback Estimation Using Deep Learning-Based Segmentation
M. J. Allen, D. Moreno-Fernández, P. Ruiz-Benito, S. W. D. Grieve, E. R. Lines
Optimizing Falsification for Learning-Based Control Systems: A Multi-Fidelity Bayesian Approach
Zahra Shahrooei, Mykel J. Kochenderfer, Ali Baheri