Songyou Peng

Papers

Total Citations

H-Index

About

Songyou Peng is a leading researcher at the intersection of computer vision and robotics, with a primary focus on neural implicit representations and depth estimation. His most impactful work introduces a groundbreaking paradigm for metric depth estimation through the novel concept of prompting depth foundation models. In his highly cited 2025 paper, "Prompting Depth Anything for 4K Resolution Accurate Metric Depth Estimation," Peng demonstrates how to leverage low-cost prompts to unlock the full potential of vision foundation models, achieving unprecedented accuracy at 4K resolution. This work has already garnered 15 citations, signaling its transformative impact on the field. Additionally, Peng has made significant contributions to the integration of Neural Radiance Fields (NeRFs) in robotics, as evidenced by his comprehensive survey papers (2024, 2025) that synthesize the state-of-the-art in using neural implicit representations for detailed 3D environment modeling. His research bridges the gap between foundational computer vision techniques and practical robotic applications, enabling more realistic and actionable 3D scene understanding. Peng’s work is essential reading for anyone interested in advancing autonomous systems through cutting-edge depth perception and neural rendering.