The RWTH/UPB/FORTH System Combination for the 4th CHiME Challenge Evaluation

By Tobias Menne, Jahn Heymann, Anastasios Alexandridis, Kazuki Irie, Albert Zeyer, Markus Kitza, Pavel Golik, Ilia Kulikov, Lukas Drude, Ralf Schlüter and 3 more. 2016

48 citations recorded by the source corpus.

Subfields: Evaluation & Benchmarks, Speech & Audio

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Evaluation & Benchmarks

1. Yoshua Bengio
2. Mirco Ravanelli
3. Maurizio Omologo
4. Reinhold Häb-Umbach
5. Jinyu Li
6. Yifan Gong
7. Philemon Brakel

In Speech & Audio

1. Yoshua Bengio
2. Mirco Ravanelli
3. Jinyu Li
4. Philemon Brakel
5. Maurizio Omologo
6. Yifan Gong
7. Reinhold Häb-Umbach
8. Yan Huang

Citations in the corpus

• Light Gated Recurrent Units for Speech Recognition (2018, methodology): “Several works have already highlighted the effectiveness of RNNs in various speech processing tasks, such as speech recognition [35]–[39], speech enhancement [40], speech separation [41], [42] as well as speech activity detection [43].”
• Improving Speech Recognition by Revising Gated Recurrent Units (2017, background): “Several works have already highlighted the effectiveness of RNNs in various speech processing tasks such as speech recognition [11, 12, 13, 14, 15], speech enhancement [16], speech separation [17, 18] as well as speech activity detection [19].”
• Improved cepstra minimum-mean-square-error noise reduction algorithm for robust speech recognition (2017, background): “However, for the deep learning based ASR systems, although beamforming multi-channel signal processing is still very helpful [15][16][17][18], it was reported that single-channel robust frontends cannot benefit deep learning models [19][20] in multi-style”
• An Investigation into the Effectiveness of Enhancement in ASR Training and Test for Chime-5 Dinner Party Transcription (2019, background): “would not only reduce the acoustic variability, it would also result in a reduction of the amount of training data by a factor of M , where M is the number of microphones [5].”
• Deep Learning for Environmentally Robust Speech Recognition (2017, unknown): “. THOSE METHODS ARE SUMMARISED AT DIFFERENT ASR PROCESSING STAGES (front-END AND joint FRONT- AND BACK-END). stage approaches typical publications advantages disadvantages front mask estimation [122]–[125] avoid relying on a DOA estimation require large-scale training data filter coefficients estimati”