Enhancing Far-Field Speech Recognition with Mixer: A Novel Data Augmentation Approach
Recent advancements in end-to-end (E2E) modeling have notably improved automatic speech recognition (ASR) systems; however, far-field speech recognition (FSR) remains challenging due to signal degradation from factors such as low signal-to-noise ratio, reverberation, and interfering sounds. This req...
保存先:
| 主要な著者: | , , , |
|---|---|
| フォーマット: | Artigo |
| 言語: | Inglês |
| 出版事項: |
MDPI AG
2025-04-01
|
| シリーズ: | Applied Sciences |
| 主題: | |
| オンライン･アクセス: | https://www.mdpi.com/2076-3417/15/7/4073 |
| タグ: |
タグなし, このレコードへの初めてのタグを付けませんか!
|