09:00 - 09:30 - Startup and coffee
09.30 - 11.20 Talks
- Steve Renals - Introduction [pdf]
- Penny Karanasou "Progress in Adaptation of DNN-based Acoustic Models" [pdf]
- Pawel Swietojanski "An overview of DNN adaptation methods for ASR Systems within NST project"
- Zhizheng Wu "An overview of DNN adaptation methods for TTS Systems within NST project"
- Peter Bell "The MGB Challenge at IEEE ASRU-2015" [pdf]
- One minute madness (introduction of demos/posters in next session) [pdf]
11.20 - 13:00 Coffee and demos/posters
Pawel Swietojanski "Learning hidden unit contributions for unsupervised speaker adaptation of neural network acoustic models"
- Yulan Liu/Penny Karanasou "An investigation into speaker informed DNN front-end for LVCSR" [pdf]
- Yulan Liu "On the relationship between speaker informed DNN training and linear DNN input normalisation" [pdf]
- Penny Karanasou "I-Vector estimation using informative priors for adaptation of deep neural networks" [pdf]
- Chunyang Wu "Multi-basis Adaptive Neural Network for Rapid Adaptation in Speech Recognition"
- Peter Bell "The UEDIN ASR Systems for the IWSLT 2014 Evaluation"
- Siva Reddy Gangireddy "Prosodically-enhanced recurrent neural network language models" [pdf]
- Yanmin Qian "Noise-aware structured DNN for robust ASR"
- Liang Lu "A study of the RNN encoder-decoder for large vocabulary ASR"
- Mortaza Doulaty "Unsupervised domain discovery" [pdf]
- Chao Zhang "A general ANN extension for HTK" [pdf]
- Liang Lu / Pawel Swietojanski / Peter Bell "Kaldi extensions at Edinburgh"
- Marcus Tomalin "Inserting filled pauses and discourse markers for disfluent speech synthesis" [pdf]
- Mirjam Wester/Gustav Eje Henter "On the subjective evaluation of synthetic speech: are we using enough listeners?" [pdf]
- Oliver Watts "Sentence-level control vectors for deep neural network speech synthesis" [pdf]
- Tom Merritt "Deep neural network context embeddings for model selection in rich-context HMM synthesis" [pdf]
- Zhizheng Wu "“Deep neural networks employing multi-task learning and stacked bottleneck features for speech synthesis”
Pierre Lanchantin "Details of the MGB Challenge data preparation" [pdf]
Pierre Lanchantin/ Christophe Veaux "Reconstructing voices within the multiple-average-voice-model framework" [pdf]
Phil Green "Browsing Oral History: English Heritage Demo"[pdf]
- Mauro Nicolao/Salil Deena "Automatic speech recognition for people with disordered speech: results from online and offline experiments" [pdf]
- Qiang Huang "User dependent interactive system using multimodal information"
- Peter Bell "GlobalVox Demo"
13:00 - 14.15 Lunch
14.15- 15.15 Talks
- Thomas Hain - Corpora: new collections and structuring diverse data
- Christophe Veaux "Voice banking and voice reconstruction" [pdf]
- Andrew Liu "Recent improvements in training, adaptation and decoding using RNNLMs"
15.15 - 15.45 Afternoon coffee
15.45 - 16:45 Discussions
- Clinical (kicked off by Stuart Cunningham)
- Media (kicked off by Phil Woodland)
- Future Challenges (kicked off by Simon King)
17.00 - 18:30 Advisory board meeting
19:30 Dinner at Emmanuel College
A few photos taken during the meeting (all photo credits: Mauro Nicolao)