WisconsinAIVision/MixNMatch
A PyTorch implementation of MixNMatch for disentangling image factors (pose, background, shape, color) and recombining them for conditional image generation.
Not currently ranked — collecting fresh signals.
star history
This repository implements MixNMatch, a CVPR 2020 generative model that learns disentangled representations of images by separating pose, background, shape, and color factors. The model uses adversarial training to encode images into separate latent spaces for each factor and a generator to synthesize new images by recombining arbitrary factor codes. Pretrained models are provided for fine-grained domains including birds, dogs, and cars.
Frequently asked
- What is WisconsinAIVision/MixNMatch?
- A PyTorch implementation of MixNMatch for disentangling image factors (pose, background, shape, color) and recombining them for conditional image generation.
- Is MixNMatch open source?
- Yes — WisconsinAIVision/MixNMatch is an open-source project tracked on heatdrop.
- What language is MixNMatch written in?
- WisconsinAIVision/MixNMatch is primarily written in Python.
- How popular is MixNMatch?
- WisconsinAIVision/MixNMatch has 966 stars on GitHub.
- Where can I find MixNMatch?
- WisconsinAIVision/MixNMatch is on GitHub at https://github.com/WisconsinAIVision/MixNMatch.