39 citations · 68 across the 4 of their papers we have counts for
4 papers
MIXPGD: Hybrid Adversarial Training for Speech Recognition Systems
Aminul Huq, Weiyi Zhang, Xiaolin Hu
Automatic speech recognition (ASR) systems based on deep neural networks are weak against adversarial perturbations. We propose mixPGD adversarial training method to improve the ro…
Speech Separation Using an Asynchronous Fully Recurrent Convolutional Neural Network
Xiaolin Hu, Kai Li, Weiyi Zhang +3
Recent advances in the design of neural network architectures, in particular those specialized in modeling sequences, have provided significant improvements in speech separation pe…
Attack on practical speaker verification system using universal adversarial perturbations
Weiyi Zhang, Shuning Zhao, Le Liu +4
In authentication scenarios, applications of practical speaker verification systems usually require a person to read a dynamic authentication text. Previous studies played an audio…
Training Stronger Baselines for Learning to Optimize
Tianlong Chen, Weiyi Zhang, Jingyang Zhou +4
Learning to optimize (L2O) has gained increasing attention since classical optimizers require laborious problem-specific design and hyperparameter tuning. However, there is a gap b…