Pengyue Hou

Papers

Total Citations

H-Index

About

Pengyue Hou is an emerging researcher specializing in adversarial machine learning and deep learning robustness, with a particular focus on developing more balanced and practical training methodologies for neural networks. Their most notable contribution, "Adversarial Fine-tune with Dynamically Regulated Adversary" (2022), addresses one of the field's most pressing challenges: the persistent trade-off between model robustness and standard performance on clean data. By introducing a dynamically regulated adversarial fine-tuning framework, Hou's work offers a promising pathway toward building models that can withstand malicious attacks without sacrificing accuracy in real-world applications such as medical diagnosis and autonomous systems — domains where both reliability and precision are critical. This work has garnered citations across the research community, reflecting its relevance to practical deployment challenges in adversarial settings. While still in the earlier stages of their research career, Hou demonstrates a clear commitment to bridging the gap between theoretical adversarial robustness and real-world applicability, positioning themselves as a thoughtful contributor to the growing field of trustworthy and reliable artificial intelligence.