Hierarchical Multi-Stage Word-to-Grapheme Named Entity Corrector for Automatic Speech Recognition.
Abhinav GargAshutosh GuptaDhananjaya GowdaShatrughan SinghChanwoo KimPublished in: INTERSPEECH (2020)
Keyphrases
- multistage
- named entities
- automatic speech recognition
- word recognition
- word error rate
- co occurrence
- noun phrases
- speech recognition
- compound words
- conversational speech
- recognition errors
- named entity recognition
- text corpus
- spontaneous speech
- speech signal
- relation extraction
- question answering
- information extraction
- hidden markov models
- text mining
- broadcast news
- named entity recognizer
- natural language processing
- dynamic programming
- unsupervised learning
- linguistic features
- text documents
- optimal policy
- out of vocabulary
- annotated corpus
- spoken document retrieval
- language model
- handwriting recognition
- statistical topic models
- word sense disambiguation
- semantic relations
- n gram
- wordnet
- word level
- natural language
- pattern recognition
- speech sounds
- keywords
- artificial intelligence