Haoxiang Wang

Papers

Total Citations

H-Index

About

Haoxiang Wang is a researcher whose work sits at the intersection of computer vision, affective computing, and deep learning, with a particular focus on facial expression recognition. His most cited paper, "Convolution by Multiplication: Accelerated Two-Stream Fourier Domain Convolutional Neural Network for Facial Expression Recognition" (2021, 27 citations), introduces an innovative approach that leverages Fourier domain operations to accelerate convolutional neural networks. By reformulating convolution as multiplication in the frequency domain, Wang’s two-stream architecture significantly improves computational efficiency while maintaining high accuracy—a critical advancement for real-time applications in psychology, human-computer interaction, and robotics. This work exemplifies his broader contributions to making deep learning models more practical and deployable. Wang’s research addresses the fundamental challenge of balancing model expressiveness with speed, a key concern for resource-constrained environments. His achievements highlight a talent for bridging theoretical signal processing techniques with applied machine learning, offering a fresh perspective on how to optimize neural networks for nuanced tasks like emotion recognition. For students and researchers, Wang’s work serves as an inspiring example of how cross-domain thinking—combining Fourier analysis with CNN design—can lead to impactful, efficient solutions in visual understanding.