Shichong Peng

Papers

Total Citations

H-Index

About

Shichong Peng is a researcher advancing the frontiers of 3D computer vision and shape analysis. His primary research focus lies in developing novel machine learning methods for multimodal shape completion, a critical problem for applications in robotics and autonomous systems where sensor data is often incomplete due to occlusion or sparsity. Peng’s major contribution is pioneering the use of Implicit Maximum Likelihood Estimation (IMLE) to address the inherent ambiguity in shape completion—rather than predicting a single deterministic output, his work generates multiple plausible, high-fidelity 3D shapes from a single partial input. This multimodal approach, detailed in his most-cited paper from 2022 (13 citations) and its 2021 precursor (6 citations), represents a significant departure from traditional deterministic methods, offering richer and more robust solutions for real-world perception tasks. By enabling systems to reason about multiple possible completions, Peng’s research directly tackles a fundamental limitation in existing shape completion pipelines, marking him as an emerging innovator in the field of generative models for 3D data.