PODIA-3D: Domain Adaptation of 3D Generative Model Across Large Domain Gap Using Pose-Preserved Text-to-Image Diffusion.
Gwanghyun KimJi Ha JangSe Young ChunPublished in: CoRR (2023)
Keyphrases
- generative model
- domain adaptation
- bayesian framework
- semi supervised
- cross domain
- image features
- input image
- probabilistic model
- image retrieval
- multiscale
- transfer learning
- image segmentation
- document classification
- image classification
- information retrieval
- target domain
- multiple sources
- prior knowledge
- domain specific
- em algorithm
- labeled data
- semi supervised learning
- topic models
- feature vectors
- feature points
- information extraction
- high resolution
- image representation
- web documents
- expectation maximization
- training data
- similarity measure
- feature representation
- principal component analysis