Juncheng Liu

Papers

Total Citations

H-Index

About

Juncheng Liu is a researcher advancing the frontiers of 3D computer vision and efficient data representation. His work focuses on tackling the critical challenge of compressing volumetric data, which is essential for enabling lightweight applications in robotics and scene understanding. Liu’s most notable contribution, "Variational Autoencoder for 3D Voxel Compression" (2020), introduces a novel deep learning framework that dramatically reduces the storage footprint of 3D voxel grids. By leveraging variational autoencoders, his method achieves high-fidelity reconstruction while maintaining compact representations, addressing a fundamental bottleneck in 3D sensing pipelines. This work has garnered 6 citations, reflecting its growing influence among researchers seeking efficient solutions for real-time 3D processing. Liu’s research sits at the intersection of generative models and geometric deep learning, offering practical pathways for deploying 3D perception in resource-constrained environments. His contributions are particularly valuable for autonomous navigation, augmented reality, and any domain where transmitting or storing large 3D scenes is prohibitive. Through his innovative approach to voxel compression, Juncheng Liu is helping to make 3D computer vision more accessible and scalable for next-generation intelligent systems.