Hao Chern
Papers
1
Total Citations
2
H-Index
1
About
Hao Chern’s research focuses on the intersection of computer vision and robotics, particularly in action-conditioned visual prediction—a critical capability for autonomous systems that must anticipate the outcomes of their movements. His most cited work, “Practical Issues of Action-Conditioned Next Image Prediction” (2018, 2 citations), provides the first systematic comparison of two influential models: Convolutional Dynamic Neural Advection (CDNA) and feedforward alternatives. By rigorously evaluating these approaches on real-world robotic tasks, Chern identified key practical challenges—such as computational efficiency and generalization across environments—that have shaped subsequent research in predictive visual modeling. While his citation count remains modest, this early-stage contribution is notable for its methodological clarity and direct relevance to embodied AI. Chern’s work serves as a valuable reference for researchers developing next-frame prediction systems for robot manipulation, offering both a benchmark and a roadmap for addressing the inherent difficulties of learning visual dynamics from action sequences. His findings continue to inform efforts in self-supervised learning and model-based control in robotics.
Research Focus
Key Achievements
Top Papers
- 1Practical Issues of Action-Conditioned Next Image Prediction2 citations · 2018