This repository contains 3 models, corresponding to the ones described in Feature-Conditioned Cascaded Video Diffusion Models for Precise Echocardiogram Synthesis. Hadrien Reynaud, Mengyun Qiao, Mischa Dombrowski, Thomas Day, Reza Razavi, Alberto Gomez, Paul Leeson and Bernhard Kainz. MICCAI 2023.
To see all the details, refer to the corresponding github repository: https://github.com/HReynaud/EchoDiffusion.
The available models are:
- 1SCM: Single Stage Cascade Model
- 2SCM: Two Stage Cascade Model
- 4SCM: Four Stage Cascade Model
All weights files contain the weights of all the diffusion model in the cascade.
To see a demo of the 1SCM, head to https://huggingface.co/spaces/HReynaud/echocardiogram-video-diffusion.
In each model folder, you will find:
config.yaml: the configuration file associated to the model. It contains the hyperparameters of the model.merged.pt: the weight file containing all the models in the cascade for that model (ex. 4 models for the 4SCM).