Login / Signup

Mitigating Closed-Model Adversarial Examples with Bayesian Neural Modeling for Enhanced End-to-End Speech Recognition.

Chao-Han Huck YangZeeshan AhmedYile GuJoseph SzurleyRoger RenLinda LiuAndreas StolckeIvan Bulyko
Published in: ICASSP (2022)
Keyphrases
  • end to end
  • speech recognition
  • probabilistic model
  • language model
  • neural network
  • machine learning
  • information retrieval
  • pattern recognition
  • em algorithm
  • posterior probability
  • speech recognizer