search

ubicomplab / rPPG-Toolbox

rPPG-Toolbox: Deep Remote PPG Toolbox (NeurIPS 2023)
https://arxiv.org/abs/2210.00716
Other
402 stars 97 forks source link

unsupervised method test-data predictions and labels #205

Closed kimyongdo1 closed 9 months ago

kimyongdo1 commented 10 months ago

hi First of all, thank you very much for your contribution to this field. This toolbox is excellent.

Currently, the neural network method can output RPPG data and display images. STEP 1: Specify the TEST.OUTPUT_SAVE_DIR field in the .yaml config file of a neural method experiment. STEP 2: Run the neural method experiment. A .pickle file containing the test-data predictions and labels will be created in TEST.OUTPUT_SAVE_DIR.

I would like to know how to output data and display images using unsupervised methods?

THANK YOU ~

yahskapar commented 10 months ago

Currently, this functionality is not available with unsupervised methods. @girishvn implemented this functionality with neural methods and he or someone else may be able to extend it to unsupervised methods sometime in the near future.

kimyongdo1 commented 10 months ago

If I generate RPPG data in an unsupervised manner, which variable should I save to obtain this data?

girishvn commented 10 months ago

Hi @kimyongdo1 you can save the BVP variable in the unsupervised_predictor.py file to save method outputs (this is the predicted BVP signal for the entirety of a sample video). I will implement the functionality to save this data in the near future.

kimyongdo1 commented 10 months ago

Got it， thank you all^^

yahskapar commented 9 months ago

I'm going to go ahead and close this issue since there doesn't appear to be any further discussion and, as @girishvn mentioned, functionality to save the data for unsupervised methods will be implemented in the near future. @kimyongdo1, feel free to re-open if needed, or create a new issue if you run into any other problem.