Jinye Peng

Papers

Total Citations

H-Index

About

Jinye Peng is a computer vision researcher whose work bridges the critical gap between 3D perception and practical augmented reality (AR) systems. His primary research areas include 6D object pose estimation, indoor localization, and multi-modal sensor fusion. Peng’s most notable contribution is his pioneering approach to learning cross-view consistent 3D keypoints for object 6D pose estimation from RGB images alone—a method that reduces reliance on expensive labeled datasets by leveraging geometric consistency across views. This work, published in 2025, has already garnered early citations for its potential to advance robotic manipulation and autonomous driving. In parallel, Peng has addressed the fundamental challenge of indoor localization by fusing WiFi signals with vision data from smart devices, a technique that directly enhances AR systems’ ability to estimate target object positions in real environments. His 2018 paper on this fusion method remains a reference point for researchers seeking robust, device-agnostic localization solutions. By tackling both the theoretical underpinnings of 3D keypoint learning and the practical constraints of real-world deployment, Peng is establishing himself as a versatile contributor to the next generation of intelligent, spatially-aware systems.