Dongxu Miao

Papers

Total Citations

H-Index

About

Dongxu Miao is a computer vision researcher whose work bridges the gap between computational photography and robotic perception. His primary research focuses on rephotography—the precise alignment of modern photographs with historical reference images—and active camera relocalization, where cameras physically return to a previous pose. Miao’s most cited work, "Fast and Reliable Computational Rephotography on Mobile Device" (2018, 15 citations), introduced a practical, automated solution for aligning images without relying on tedious human judgment, enabling applications in historical documentation, environmental monitoring, and change detection. His second major contribution, "Active Camera Relocalization with RGBD Camera from a Single 2D Image" (2018, 8 citations), advanced robotic perception by allowing cameras to dynamically reposition themselves using depth information, supporting automated picking, stowing, and fine-grained change detection. These works demonstrate Miao’s talent for making complex computer vision techniques both fast and reliable for real-world deployment. His research is particularly notable for its focus on mobile and robotic platforms, where computational efficiency is critical. For students and researchers, Miao’s work exemplifies how combining geometric computer vision with practical engineering can solve challenging problems in cultural heritage, robotics, and environmental science.