Adversarial Teacher-Student Learning for Unsupervised Domain Adaptation

Abstract

The teacher-student (T/S) learning has been shown effective in unsupervised domain adaptation [1]. It is a form of transfer learning, not in terms of the transfer of recognition decisions, but the knowledge of posteriori probabilities in the source domain as evaluated by the teacher model. It learns to handle the speaker and environment variability inherent in and restricted to the speech signal in the target domain without proactively addressing the robustness to other likely conditions. Performance degradation may thus ensue. In this work, we advance T/S learning by proposing adversarial T/S learning to explicitly achieve condition-robust unsupervised domain adaptation. In this method, a student acoustic model and a condition classifier are jointly optimized to minimize the Kullback-Leibler divergence between the output distributions of the teacher and student models, and simultaneously, to min-maximize the condition classification loss. A condition-invariant deep feature is learned in the adapted student model through this procedure. We further propose multi-factorial adversarial T/S learning which suppresses condition variabilities caused by multiple factors simultaneously. Evaluated with the noisy CHiME-3 test set, the proposed methods achieve relative word error rate improvements of 44.60% and 5.38%, respectively, over a clean source model and a strong T/S learning baseline model.

5 pages, 1 figure, ICASSP 2018

References in corpus (6)

• Unsupervised Domain Adaptation by Backpropagation
• Domain Separation Networks
• Speaker-Invariant Training via Adversarial Learning
• Deep Long Short-Term Memory Adaptive Beamforming Networks For Multichannel Robust Speech Recognition
• Unsupervised Adaptation with Domain Separation Networks for Robust Speech Recognition
• Invariant Representations for Noisy Speech Recognition

Cited by in corpus (25)

• Speaker-Invariant Training via Adversarial Learning
• Conditional Teacher-Student Learning
• Adaptation Algorithms for Neural Network-Based Speech Recognition: An Overview
• Adversarial Speaker Verification
• Cycle-Consistent Speech Enhancement
• KD3A: Unsupervised Multi-Source Decentralized Domain Adaptation via Knowledge Distillation
• Speaker Adaptation for Attention-Based End-to-End Speech Recognition
• Adversarial Feature-Mapping for Speech Enhancement
• Adversarial Speaker Adaptation
• Semi-Supervised Speech Recognition via Local Prior Matching
• Narrowing the Gap: Improved Detector Training with Noisy Location Annotations
• Does Knowledge Distillation Really Work?
• Attentive Adversarial Learning for Domain-Invariant Training
• Recent Progresses in Deep Learning based Acoustic Models (Updated)
• Adversarial Imitation Learning with Trajectorial Augmentation and Correction
• Active Voice Authentication
• Internal Language Model Training for Domain-Adaptive End-to-End Speech Recognition
• Domain Adaptation via Teacher-Student Learning for End-to-End Speech Recognition
• Interpretable Foreground Object Search As Knowledge Distillation
• Unsupervised Domain Expansion for Visual Categorization
• Teacher-Student Consistency For Multi-Source Domain Adaptation
• Fine-grained Knowledge Fusion for Sequence Labeling Domain Adaptation
• Minimum Word Error Rate Training with Language Model Fusion for End-to-End Speech Recognition
• Deep Least Squares Alignment for Unsupervised Domain Adaptation
• Multimodal and Multi-view Models for Emotion Recognition