• Home

• People

• Publications

• Research

• Data

• Courses

• Software

• Meetings

SST Group Meetings, Recent

• Thursday, May 2: Midwest Speech and Language Days
• Thursday, Apr 18, 11:00-12:00, Beckman 3169:
  • Jennifer Zhang presents a summary of spectral moments for word-final fricative classifiation
  • Liming Wang presents Kamper, Livescu and Goldwater, An embeddged segmental k-means model for unsupervised segmentation and clustering of speech
  • Mahir Morshed presents Emre Yılmaz, Jelske Dijkstra, Hans Van de Velde, Frederik Kampstra, Jouke Algra, Henk van den Heuvel, David Van Leeuwen, Longitudinal Speaker Clustering and Verification Corpus with Code-Switching Frisian-Dutch Speech
  • Yda Hoffer-Sohn presents Chen, Yilmaz, Trmal, Povey and Khudanpur, Using Proxies for OOV Keywords in the Keyword Search Task
• Thursday, Apr 4, 11:00-12:00, Beckman 3169: Yijia Xu presented WFSTs for Automatic Speech Recognition.
• Thursday, Mar 14, 11:00-12:00, Beckman 3169:
  Liming Wang presents A Spectral Algorithm for Learning Class-Based n-gram Models of Natural Language
  Jialu Li presents Language Independent End-to-End Architecture for Joint Language Identification and Speech Recognition
• Thursday, Mar 7, 11:00-12:00, Beckman 3169:
  How can speech technologies support learners to improve their skills of speaking, listening, conversation and more?
  Nobuaki Minematsu, Tokyo University
  Slides and Papers
• Thursday, Feb 28, 11:00-12:00, Beckman 3169:
  Leda Sari presents Hybrid CTC/Attention Architecture for End-to-End Speech Recognition, by Shinji Watanabe, Suyoun Kim, John Hershey, and Tomooki Hayashi,
  Ali Abavisani presents Leigh Lisker, Is it VOT or a first-formant transition detector?, JASA 57(6):1547-1551, 1975
• Thursday, Feb 21, 11:00-12:00, Beckman 3169: Raymond Yeh presents a Tutorial on Variational Autoencoder and its Gradient Estimators.
• Thursday, Feb 14, 11:00-12:00, Beckman 3169:
  • Mahir Morshed presents Sequence-based multi-lingual low resource speech recognition by Siddharth Dalmia, Ramon Sanabria, Florian Metze, and Alan W. Black, ICASSP 2018
  • Yda Hoffer-Sohn presents Bootstrapping Text-to-Speech for Speech Processing in Languages Without an Orthography by Sunayana Sitaram, Sukhada Palkar, Yun-Nung Chen, Alok Parlik, and Alan W. Black, ICASSP 2013
  • Heting Gao presents LanczosNet: Multi-Scale Deep Graph Convolutional Networks
• Thursday, Feb 7, 11:00-12:00, ECEB 3032:
  • Jennifer Zhang presents Assessing the distinctiveness of phonological features in word recognition: Prelexical and lexical influences by Alexander Martin and Sharon Peperkamp, J Phonetics 62:1-11, 2017
  • Liming Wang presents Direct Optimization of Ranking Measures by Quoc V. Le and Alex J. Smola, arxiv preprint, 2007 (slides)
• Thursday, Jan 31, 11:00-12:00, ECEB 3032:
  • Leda Sari presents Analyzing Hidden Representations in End-to-End Automatic Speech Recognition Systems, Yonatan Belinkov and James Glass, NIPS 2017 (slides)
  • Jialu Li presents Robust Speech Recognition Using Generative Adversarial Networks by Anuroop Sriram, Heewoo Jun, Yashesh Gaur, and Sanjeev Satheesh, ICASSP 2018, 5639-5643
  • Yijia Xu presents The Kaldi OpenKWS System: Improving Low Resource Keyword Search, by Jan Trmal et al., Interspeech 2017
  • Guohao Dou presents A Least Squares Formulation for Canonical Correlation Analysis, Liang Sun, Shuiwant Ji, and Jieping Ye, ICML 20008 (two-page note)
• Monday, 12/3, 14:00, ECEB 3032: Guohao Dou, pytorch pre-processing (PDF), see also his github.
• Friday, 11/16, 14:00, Beckman 4169: Jingning Tang presents Learning Type-Aware Embeddings for Fashion Compatibility, Mariya I. Vasileva, Bryan A. Plummer, Krishna Dusad, Shreya Rajpal, Ranjitha Kumar, and David Forsyth, ECCV 2018.
• Monday, 11/12, 16:00, ECEB 3032: 5-minute status updates from each member
• Friday, 11/9, 14:00, Beckman 4169: Heting Gao discusses Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering and Semi-Supervised Classification with Graph Convolutional Networks. Heting's slides are here.
• Monday, 10/29, 16:00, ECEB 3032: Liming Wang, TensorFlow for Image-to-Speech and Speech-to-Image retrieval.
• Friday, 10/26, 14:00, Beckman 4169: Guohao Dou discusses On Deep Multi-View Representation Learning, Weiran Wang, Raman Arora, Karen Livescu and Jeff Bilmes, ICML 2015
• Monday, 10/22, 16:00, ECEB 3032: Short status reports: Claire, Leda, Liming, Mahir, Yda.
• Friday, 10/19, 14:00, Beckman 4169: Liming Wang, Hessian-free Optimization for Learning Deep Multidimensional Recurrent Neural Networks, Minhyung Cho, Chandra Shekhar Dhir, and Jaehyung Lee. NIPS 2015.
• Friday, 10/12, 14:00, Beckman 4169: Leda Sari discusses X-vectors: Robust DNN Embeddings for Speaker Recognition. David Snyder, Daniel Garcia-Romero, Gregory Sell, Daniel Povey, Sanjeev Khudanpur. ICASSP 2018.
• Monday, 10/8, 16:00, ECEB 3032: Neural random fields with applications to modeling of languages and images, Zhijian Ou, Tsinghua University
• Friday, 10/5, 14:00, Beckman 4169: Feng Li discusses Wave-U-Net: A Multi-Scale Neural Network for End-to-End Audio Source Separation Daniel Stoller, Sebastian Ewert, Simon Dixon, ISMIR 2018.
• Monday, 10/1, 16:00, ECEB 3032: Leda Sari, Kaldi tutorial. Audio-visual speech recognition.
• Friday, 9/28, 14:00, Beckman 4169: Raymond Yeh discusses Phased LSTM: Accelerating Recurrent Network Training for Long or Event-based Sequences, Daniel Neil, Michael Pfeiffer, and Shih-Chii Liu, NIPS 2016.
• Monday, 9/24, 16:00, ECEB 3032: Jialu Li, "Learning a Better Phone Set by Clustering the Embeddings of Listen, Attend and Spell."
• Monday, 9/19 (experiments): each group member outlined a planned experiment. Topics included sequence-to-sequence models (for machine translation, and a clustering of acoustic encoding vectors in listen-attend-spell), speaker change detection using siamese networks, image-to-speech, NMF-DNN hybrids (for speech enhancement, and for infant cry classification), and dialect variation (in German and Bengali).
• Monday, 8/27 (experiments): comparison of available toolkits, focusing on Kaldi, TensorFlow, and XNMT. Brief tutorial in the purposes and uses of WFSTs.
• Monday, 8/27 (theory): planning meeting.

Archives: 1996-2018

2018S | 2017 | 2016F | 2016S | 2015F | 2015S | 2014F | 2014S | 2013F | 1996-2010

Midwest Speech and Language Days: 2018, 2017, 2016, 2015, 2014, 2013, 2012, 2011, 2010, 2009

Focal Point Projects: Bilingualism, Speech Production

Speech Prosody 2010