BanSpeech: A Multi-Domain Bangla Speech Recognition Benchmark Toward Robust Performance in Challenging Conditions.
Ahnaf Mozib SaminM. Humayon KobirMd. Mushtaq Shahriyar RafeeM. Firoz AhmedMehedi HasanPartha GhoshShafkat KibriaM. Shahidur RahmanPublished in: IEEE Access (2024)
Keyphrases
- speech recognition
- multi domain
- noisy environments
- hidden markov models
- speech synthesis
- speech signal
- automatic speech recognition
- speech processing
- cross domain
- language model
- pattern recognition
- speech recognizer
- speech recognition technology
- speech recognition systems
- domain specific
- handwriting recognition
- speaker identification
- real world
- audio visual speech recognition
- heterogeneous networks
- text to speech
- transfer learning
- co occurrence
- neural network