Towards hate speech detection in low-resource languages: Comparing ASR to acoustic word embeddings on Wolof and Swahili.
Christiaan JacobsNathanaël Carraz RakotonirinaEverlyn Asiko ChimotoBruce A. BassettHerman KamperPublished in: INTERSPEECH (2023)
Keyphrases
- automatic speech recognition
- speech recognition
- speech recognition systems
- speech sounds
- speech recognizers
- spontaneous speech
- word error rate
- spoken language
- speech recognizer
- speech signal
- prosodic features
- acoustic features
- compound words
- recognition errors
- english text
- noisy environments
- speech segments
- conversational speech
- word recognition
- speech retrieval
- language model
- hidden markov models
- broadcast news
- speaker independent
- speaker identification
- language independent
- speech synthesis
- spoken document retrieval
- spoken words
- language identification
- statistical machine translation
- target language
- language specific
- acoustic models
- information retrieval
- co occurrence
- n gram
- false positives
- multi lingual
- vector space
- speech corpus
- cross lingual
- handwriting recognition
- low dimensional
- dimensionality reduction
- vocal tract