S. Mishra

Papers

Total Citations

H-Index

About

S. Mishra is a researcher pushing the boundaries of 3D computer vision, with a focused expertise in shape completion—the critical task of reconstructing full 3D objects from partial, often noisy input data like sensor scans. Their major contribution lies in pioneering multimodal approaches to this ill-posed problem, specifically through the application of Implicit Maximum Likelihood Estimation (IMLE). Mishra’s work directly addresses a core limitation of prior methods: their tendency to produce only a single, deterministic output. By introducing IMLE into shape completion, Mishra enables models to generate multiple, plausible completions from the same partial input, capturing the inherent ambiguity of real-world data. This innovation is vital for applications in robotics and autonomous systems, where occlusion and sensor sparsity are common. Their most-cited paper, “Multimodal Shape Completion via Implicit Maximum Likelihood Estimation” (2022), has garnered 13 citations, while a related 2021 work has 6, establishing a growing footprint in the field. Mishra’s research is notable for its practical focus on handling the uncertainty of real-world scans, offering a more robust and flexible solution compared to deterministic baselines. For students and researchers, Mishra’s work represents a key step toward making 3D perception systems more reliable and human-like in their ability to infer missing information.