PM3: DNN-based ASR |
| Session Type: Poster |
| Poster Time: Monday, December 8, 16:00 - 17:30 |
| Location: Emerald 4-6 |
| Session Chair: Ozlem Kalinli, Sony Playstation
|
| |
| PM3.101: DEEP ORDER STATISTIC NETWORKS |
| Steven J. Rennie; IBM Research, United States |
| Vaibhava Goel; IBM Research, United States |
| Samuel Thomas; IBM Research, United States |
| |
| PM3.102: IMPROVING DEEP NEURAL NETWORKS USING STATE PROJECTION VECTORS OF SUBSPACE GAUSSIAN MIXTURE MODEL AS FEATURES |
| Murali Karthick B; Indian Institute of Technology Madras, India |
| S Umesh; Indian Institute of Technology Madras, India |
| |
| PM3.103: VOCAL TRACT LENGTH NORMALISATION APPROACHES TO DNN-BASED CHILDREN'S AND ADULTS' SPEECH RECOGNITION |
| Romain Serizel; FBK- Fondazione Bruno Kessler, Italy |
| Diego Giuliani; FBK- Fondazione Bruno Kessler, Italy |
| |
| PM3.104: SEMI-SUPERVISED DNN TRAINING IN MEETING RECOGNITION |
| Pengyuan Zhang; Chinese Academy of Sciences, China |
| Yulan Liu; University of Sheffield, United Kingdom |
| Thomas Hain; University of Sheffield, United Kingdom |
| |
| PM3.105: TIKHONOV REGULARIZATION FOR DEEP NEURAL NETWORK ACOUSTIC MODELING |
| Jen-Tzung Chien; National Chiao Tung University, Taiwan |
| Tsai-Wei Lu; National Chiao Tung University, Taiwan |
| |
| PM3.106: SPEAKER ADAPTATION OF DEEP NEURAL NETWORKS USING A HIERARCHY OF OUTPUT LAYERS |
| Ryan Price; Tokyo Institute of Technology, Japan |
| Ken-ichi Iso; Yahoo Japan Corporation, Japan |
| Koichi Shinoda; Tokyo Institute of Technology, Japan |
| |
| PM3.107: ANNEALED DROPOUT TRAINING OF DEEP NETWORKS |
| Steven J. Rennie; IBM Research, United States |
| Vaibhava Goel; IBM Research, United States |
| Samuel Thomas; IBM Research, United States |
| |
| PM3.108: IMPROVEMENTS TO SPEAKER ADAPTIVE TRAINING OF DEEP NEURAL NETWORKS |
| Yajie Miao; Carnegie Mellon University, United States |
| Lu Jiang; Carnegie Mellon University, United States |
| Hao Zhang; Carnegie Mellon University, United States |
| Florian Metze; Carnegie Mellon University, United States |
| |
| PM3.109: LEARNING HIDDEN UNIT CONTRIBUTIONS FOR UNSUPERVISED SPEAKER ADAPTATION OF NEURAL NETWORK ACOUSTIC MODELS |
| Pawel Swietojanski; University of Edinburgh, United Kingdom |
| Steve Renals; University of Edinburgh, United Kingdom |
| |
| PM3.110: GRAPH-BASED SEMI-SUPERVISED ACOUSTIC MODELING IN DNN-BASED SPEECH RECOGNITION |
| Yuzong Liu; University of Washington, United States |
| Katrin Kirchhoff; University of Washington, United States |
| |
| PM3.111: A DISTRIBUTED ARCHITECTURE FOR FAST SGD SEQUENCE DISCRIMINATIVE TRAINING OF DNN ACOUSTIC MODELS |
| George Saon; IBM, United States |
| |