Residual-Guided Personalized Speech Synthesis based on Face Image.
Jianrong WangZixuan WangXiaosheng HuXuewei LiQiang FangLi LiuPublished in: ICASSP (2022)
Keyphrases
- speech synthesis
- face images
- face recognition
- speech recognition
- text to speech
- human faces
- principal component analysis
- low resolution
- facial expressions
- feature vectors
- illumination variations
- age estimation
- high resolution
- feature extraction
- training set
- face representation and recognition
- face recognition systems
- facial features
- face verification
- pose variations
- feature points
- cmu pie
- human face recognition
- low resolution face images
- neural network
- face databases
- image set
- pattern recognition
- face recognition algorithms
- face representation
- recognition rate
- training data
- illumination conditions
- face model
- lighting conditions
- hidden markov models
- feature space
- orl face database
- computer vision
- real time