The RoyalFlush Automatic Speech Diarization and Recognition System for In-Car Multi-Channel Automatic Speech Recognition Challenge.
Jingguang TianShuaishuai YeShunfei ChenYang XiangZhaohui YinXinhui HuXinkang XuPublished in: CoRR (2024)
Keyphrases
- automatic speech recognition
- multi channel
- speech recognition
- speech signal
- broadcast news
- word error rate
- hidden markov models
- speaker identification
- speaker independent
- single channel
- conversational speech
- spontaneous speech
- speech corpus
- noisy environments
- lesion segmentation
- acoustic features
- spoken words
- speaker diarization
- speech retrieval
- recognition errors
- speech recognizers
- speech sounds
- word recognition
- speech segments
- speech recognizer
- acoustic models
- speaker adaptation
- vocal tract
- handwriting recognition
- pattern recognition
- feature extraction
- speech enhancement
- neural network
- multi modal
- phoneme recognition
- image processing
- speech activity detection
- machine learning