Low-Latency Speech Enhancement via Speech Token Generation.
Huaying XueXiulian PengYan LuPublished in: ICASSP (2024)
Keyphrases
- speech enhancement
- low latency
- speech signal
- noisy environments
- noise reduction
- signal to noise ratio
- single channel
- vocal tract
- high speed
- high throughput
- real time
- noisy speech
- linear prediction
- background noise
- highly efficient
- virtual machine
- speech recognition
- additive noise
- multi channel
- smoothing algorithm
- automatic speech recognition
- wiener filter
- stream processing
- speech synthesis
- low cost
- hidden markov models
- data structure
- video sequences