SUPERB-SG: Enhanced Speech processing Universal PERformance Benchmark for Semantic and Generative Capabilities.
Hsiang-Sheng TsaiHeng-Jui ChangWen-Chin HuangZili HuangKushal LakhotiaShu-Wen YangShuyan DongAndy T. LiuCheng-I Jeff LaiJiatong ShiXuankai ChangPhil HallHsuan-Jui ChenShang-Wen LiShinji WatanabeAbdelrahman MohamedHung-yi LeePublished in: CoRR (2022)
Keyphrases
- speech processing
- speech recognition
- signal processing
- natural language processing
- speaker identification
- multimedia systems
- natural language
- generative model
- machine learning
- artificial intelligence
- high level
- similarity measure
- english text
- multi modal
- variable length
- speech signal
- maximum likelihood
- semantic roles
- information retrieval systems
- pattern recognition
- multimedia
- information retrieval