Serena Mou
Papers
1
Total Citations
2
H-Index
1
About
Serena Mou is a robotics researcher whose work bridges the gap between simulation and real-world deployment through modular deep reinforcement learning (RL). Her key contributions lie in developing frameworks that disentangle perception, decision-making, and low-level control, enabling more efficient and transferable robotic learning. In her notable 2018 paper, "Zero-shot Sim-to-Real Transfer with Modular Priors," Mou introduced a modular architecture that incorporates prior knowledge into RL systems, dramatically reducing the need for joint end-to-end training from sparse rewards and high-dimensional inputs. This approach addresses a critical bottleneck in robotics—the prohibitively long training times required for sim-to-real transfer—by allowing individual modules to be pretrained and fine-tuned independently. While her citation count is modest, her work has been influential in advancing modular RL methods that prioritize sample efficiency and real-world applicability. Mou’s research continues to shape how robots learn complex tasks by breaking down the learning process into manageable, transferable components, making her a rising voice in the field of embodied AI and robotic autonomy.
Research Focus
Key Achievements
Top Papers
- 1Zero-shot Sim-to-Real Transfer with Modular Priors.2 citations · 2018