• Publications
  • Influence
Extractive summarization of meeting recordings
TLDR
This work investigates how the summarization results might deteriorate when carried out on ASR output as opposed to manual transcripts, and contrasts feature-based approaches using prosodic and lexical features with maximal marginal relevance and latent semantic analysis approaches to summarization. Expand
Learning hidden unit contributions for unsupervised speaker adaptation of neural network acoustic models
This paper proposes a simple yet effective model-based neural network speaker adaptation technique that learns speaker-specific hidden unit contributions given adaptation data, without requiring anyExpand
WSJCAMO: a British English speech corpus for large vocabulary continuous speech recognition
TLDR
The motivation for the corpus, the processes undertaken in its construction and the utilities needed as support tools are described, and comparative results on these tasks for British and American English are concluded. Expand
Automatic Dialect Detection in Arabic Broadcast Speech
TLDR
This work investigates different approaches for dialect identification in Arabic broadcast speech, using phonetic, lexical features obtained from a speech recognition system, and acoustic features using the i-vector framework, and combined these features using a multi-class Support Vector Machine (SVM). Expand
Evaluating Automatic Summaries of Meeting Recordings
TLDR
The research below explores schemes for evaluating automatic summaries of business meetings, using the ICSI Meeting Corpus, with a central interest being whether or not the two types of evaluations correlate with each other. Expand
Learning Hidden Unit Contributions for Unsupervised Acoustic Model Adaptation
This work presents a broad study on the adaptation of neural network acoustic models by means of learning hidden unit contributions (LHUC) -- a method that linearly re-combines hidden units in aExpand
Speaker-adaptation for hybrid HMM-ANN continuous speech recognition system
TLDR
These techniques are applied to a well trained, speaker-independent, hybrid HMM-ANN system and the recognizer parameters are adapted to a new speaker through o -line procedures and show that speaker-adaptation within the hybrid framework can substantially improve system performance. Expand
Dynamic Evaluation of Neural Sequence Models
TLDR
Dynamic evaluation improves the state-of-the-art word-level perplexities on the Penn Treebank and WikiText-2 datasets to 51.1 and 44.3 respectively and character-level cross-entropies on the text8 and Hutter Prize datasets to 1.19 bits/char and 1.08bits/char respectively. Expand
Multiplicative LSTM for sequence modelling
TLDR
It is demonstrated empirically that mLSTM outperforms standard LSTM and its deep variants for a range of character level language modelling tasks, and is argued makes it more expressive for autoregressive density estimation. Expand
...
1
2
3
4
5
...