Linxue Bai

Research output

Research output per year 2015 2015 2016 2018 2018

• 7 Conference contribution
• 1 Paper

Research output per year

Research output per year

Search results

• 2018

  Exploring how phone classification neural networks learn phonetic information by visualising and interpreting bottleneck features

  Bai, L., Weber, P., Jancovic, P. & Russell, M., 3 Sept 2018, Proceedings of Interspeech 2018. Hyderabad, India: ISCA, p. 1472-1476 5 p. (Interspeech; vol. 2018).

  Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

  Open Access

  File

  • Speech analysis 100%
  • Neural networks 51%
  • Discriminant analysis 36%
  • Visualization 26%
  • Speech recognition 16%
  5 Citations (Scopus)

  388 Downloads (Pure)

• Phone recognition using a non-linear manifold with broad phone class dependent DNNs

  Qian, M., Bai, L., Jancovic, P. & Russell, M., 3 Sept 2018, Proceedings of Interspeech 2018. Hyderabad, India: ISCA, p. 3753-3757 5 p. (Interspeech; vol. 2018).

  Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

  Open Access

  File

  • Speech analysis 100%
  • Speech recognition 80%
  • Decoding 72%
  1 Citation (Scopus)

  213 Downloads (Pure)
• 2017

  Phone classification using a non-linear manifold with broad phone class dependent DNNs

  Bai, L., Jancovic, P., Russell, M., Weber, P. & Houghton, S., 24 Aug 2017, Proceedings of Interspeech 2017. 5 p.

  Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

  Open Access

  File

  • Deep neural networks 100%
  • Fusion reactions 41%
  • Speech analysis 30%
  • Speech recognition 24%
  • Acoustics 21%
  3 Citations (Scopus)

  218 Downloads (Pure)
• 2016

  Interpretation of low dimensional neural network bottleneck features in terms of human perception and production

  Weber, P., Bai, L., Russell, M., Jancovic, P. & Houghton, S., 10 Jun 2016, (Accepted/In press) Proceedings of Interspeech. ISCA, 5 p.

  Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

  • Neural networks 100%
  • Speech analysis 76%
  • Acoustics 26%
  • Dynamic models 26%
  • Trajectories 22%
  4 Citations (Scopus)

• Progress on Phoneme Recognition with a Continuous-State HMM

  Weber, P., Bai, L., Houghton, S., Jancovic, P. & Russell, M., 19 May 2016, (E-pub ahead of print) 2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP): Proceedings. IEEE Xplore, 5 p.

  Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

  • Network layers 100%
  • Speech recognition 84%
  • Statistical Models 81%
  • Neural networks 54%
  3 Citations (Scopus)
• 2015

  Analysis of a Low-Dimensional Bottleneck Neural Network Representation of Speech for Modelling Speech Dynamics

  Bai, L., Jancovic, P., Russell, M. & Weber, P., 2015, Proceedings of Interspeech. ISCA, p. 583-587 4 p.

  Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

  12 Citations (Scopus)

• Consonant Recognition with Continuous-Sate Hidden Markov Models and Perceptually-Motivated Features

  Weber, P., Bai, L., Houghton, S., Jancovic, P. & Russell, M., 2015, Proceedings of Interspeech. ISCA, p. 1893-1897 4 p.

  Research output: Chapter in Book/Report/Conference proceeding › Conference contribution