Papers
7
Total Citations
98
H-Index
3
About
Yiming Hu is a researcher specializing in deep learning, neural network optimization, and autonomous mobile robotics, with a particular focus on bridging the gap between advanced AI models and real-world deployment constraints. His most influential work, "A Novel Channel Pruning Method for Deep Neural Network Compression" (2018, 67 citations), addressed one of the field's pressing challenges: making computationally intensive deep neural networks viable for resource-constrained embedded devices such as smartphones and mobile robots. This contribution established him as a meaningful voice in the model compression community. Hu's research portfolio has since evolved toward autonomous navigation, where he investigates deep reinforcement learning approaches for mapless robot navigation in unknown environments. His work tackles persistent challenges such as local optima avoidance, collision-free path planning, and dynamic obstacle avoidance, with notable contributions including motion-prediction-based DRL methods and maximum entropy learning frameworks for autonomous agents. His generative modeling work on difference-guided GANs for future frame prediction further demonstrates his breadth across computer vision and sequential decision-making. With a growing body of publications accumulating nearly 100 total citations, Hu's research offers practical, learning-driven solutions that push mobile robotics toward greater autonomy and real-world reliability.
Research Focus
Key Achievements
Top Papers
- 1A novel channel pruning method for deep neural network compression67 citations · 2018
- 2
- 3
- 4
- 5
- 6Autonomous Mapless Navigation via Maximum Entropy Learning2 citations · 2023
- 7