A Low-Power Hardware Accelerator of MFCC Extraction for Keyword Spotting in 22nm FDSOI.
Liyuan GuoMatthias JobstJohannes PartzschStefan ScholzeAndreas DixiusMatthias LohrmannSeyed Mohammad Ali ZeinolabedinChristian MayrPublished in: AICAS (2023)
Keyphrases
- low power
- keyword spotting
- low cost
- speech recognition
- single chip
- cmos technology
- digital signal processing
- vlsi architecture
- power consumption
- low power consumption
- high speed
- power reduction
- hardware and software
- speech processing
- nm technology
- hidden markov models
- signal processor
- speech signal
- gate array
- image sensor
- real time
- field programmable gate array
- embedded systems
- logic circuits
- hardware implementation
- language model
- printed documents
- feature extraction
- speaker identification
- noisy environments
- handwriting recognition
- signal processing
- mixed signal
- automatic speech recognition
- power dissipation
- computing systems
- feature set
- pattern recognition
- artificial intelligence
- information retrieval
- machine learning