Latents2Segments: Disentangling the Latent Space of Generative Models for Semantic Segmentation of Face Images.
Snehal Singh TomarA. N. RajagopalanPublished in: CoRR (2022)
Keyphrases
- generative model
- face images
- latent space
- object categories
- conditional random fields
- superpixels
- face recognition
- input image
- probabilistic model
- human faces
- principal component analysis
- facial expressions
- low resolution
- em algorithm
- bayesian framework
- image set
- high resolution
- test images
- feature vectors
- prior knowledge
- object classes
- training samples
- expectation maximization
- topic models
- training set
- multiscale
- semi supervised
- feature extraction
- data sets
- feature points
- super resolution
- latent variables
- object segmentation
- computer vision
- object class
- scene classification
- dimensionality reduction
- machine learning