Interspeech 2018 2018

End-to-end speech recognition using lattice-free MMI

Interspeech 2018 2018 kaldi-asr/kaldi

We present our work on end-to-end training of acoustic models using the lattice-free maximum mutual information (LF-MMI) objective function in the context of hidden Markov models.

END-TO-END SPEECH RECOGNITION SPEECH RECOGNITION

Semi-Orthogonal Low-Rank Matrix Factorization for Deep Neural Networks

Interspeech 2018 2018 kaldi-asr/kaldi

Time Delay Neural Networks (TDNNs), also known as onedimensional Convolutional Neural Networks (1-d CNNs), are an efficient and well-performing neural network architecture for speech recognition.

SPEECH RECOGNITION