Towards a Competitive End-to-End Speech Recognition for CHiME-6 Dinner Party Transcription

By Andrei Andrusenko, Aleksandr Laptev, Ivan Medennikov. 2020

18 citations recorded by the source corpus.

While end-to-end ASR systems have proven competitive with the conventional hybrid approach, they are prone to accuracy degradation when it comes to noisy and low-resource conditions. In this paper, we argue that, even in such difficult cases, some end-to-end approaches show performance close to the hybrid baseline. To demonstrate this, we use the CHiME-6 Challenge data as an example of challenging environments and noisy conditions of everyday speech. We experimentally compare and analyze CTC-Attention versus RNN-Transducer approaches along with RNN versus Transformer architectures. We also provide a comparison of acoustic features and speech enhancements. Besides, we evaluate the effectiveness of neural network language models for hypothesis re-scoring in low-resource conditions. Our best end-to-end model based on RNN-Transducer, together with improved beam search, reaches quality by only 3.8% WER abs. worse than the LF-MMI TDNN-F CHiME-6 Challenge baseline. With the Guided Source Separation based training data augmentation, this approach outperforms the hybrid baseline system by 2.7% WER abs. and the end-to-end system best known before by 25.7% WER abs.

Subfields: Artificial Intelligence, Speech & Audio, Signal Processing

How researchers see this paper

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

In Artificial Intelligence

1. Jinyu Li
2. Dong Yu
3. Li Deng
4. Yifan Gong
5. Zhuo Chen
6. Chin-Hui Lee
7. Zhong Meng
8. Sabato Marco Siniscalchi

In Speech & Audio

1. Jinyu Li

In Signal Processing

1. Jinyu Li
2. Dong Yu
3. Yifan Gong
4. Li Deng
5. Zhong Meng
6. Zhuo Chen
7. Chin-Hui Lee
8. Sabato Marco Siniscalchi

Citations in the corpus

• SpeechStew: Simply Mix All Available Speech Recognition Data to Train One Large Neural Network (2021, neutral): “For example, the best previously published end-to-end model achieved 49.0% WER on the CHiME-6 dev set [8], while the best HMM model achieves 36.9% WER [9].”
• You Do Not Need More Data: Improving End-To-End Speech Recognition by Text-To-Speech Data Augmentation (2020, neutral): “But for lowresource tasks, end-to-end recognition quality is far behind from the hybrid models [2], [3].”
• Recent Advances in End-to-End Automatic Speech Recognition (2021, background): “However, there is not too much activity in addressing such mismatch [4, 217].”