SPECPATCH: Human-In-The-Loop Adversarial Audio Spectrogram Patch Attack on Speech Recognition.
Hanqing GuoYuanda WangNikolay IvanovLi XiaoQiben YanPublished in: CCS (2022)
Keyphrases
- speech recognition
- speech signal
- automatic speech recognition
- speaker identification
- speech processing
- speech recognition technology
- hidden markov models
- speech recognizer
- broadcast news
- audio visual speech recognition
- pattern recognition
- speech synthesis
- language model
- cepstral coefficients
- speech recognition systems
- noisy environments
- acoustic features
- multimedia
- multi stream
- signal processing
- speech recognizers
- speaker independent
- voice activity detection
- speaker diarization
- speaker recognition
- audio visual
- probabilistic model
- computer vision
- mel frequency cepstral coefficients
- non stationary
- speech recognition errors
- feature extraction
- neural network