Thank you for releasing the UniCardio codebase and for the impressive work on unified cardiovascular signal modeling. We have been carefully studying the repository, especially the downstream AF task under down_stream_code/AF, and find it highly relevant to our ongoing research.
We would like to ask whether you plan to (or are able to) release any pretrained downstream weights for AF detection. Specifically, we are interested in:
- Pretrained downstream AF classification weights used in your experiments, only the model (VggECG)
- Whether any PPG-based AF detection (e.g., PPG → AF) models or checkpoints were trained internally, even if not included in the current release
Thank you again for making this work publicly available. We truly appreciate your effort and would be happy to acknowledge or cite appropriately if any guidance or resources can be shared.
Thank you for releasing the UniCardio codebase and for the impressive work on unified cardiovascular signal modeling. We have been carefully studying the repository, especially the downstream AF task under
down_stream_code/AF, and find it highly relevant to our ongoing research.We would like to ask whether you plan to (or are able to) release any pretrained downstream weights for AF detection. Specifically, we are interested in:
Thank you again for making this work publicly available. We truly appreciate your effort and would be happy to acknowledge or cite appropriately if any guidance or resources can be shared.