Marek

Hi Young, `frameFraction` refers to the fraction of the input image that will be the "frame", which is the area outside of the landamarks of `initLandmarks`. If you set this...

Hi, you can easily modify the CameraDemo.py script to load a single image instead of a camera frame. Marek

Hi, It basically implements appendix D from this very old paper: http://www2.compute.dtu.dk/courses/02511/docs/asm_overview.pdf Marek

Hi, The scripts for creating the datasets are the same as in the paper, the pre-trained DNNs should give results exactly the same as reported in the paper. Which DNNs...

Hi, like I said before, the results should be identical to the ones in the paper. Which DNN did you use and what results do you get on the other...

Hi, The bounding boxes are from the 300W dataset. Marek

Hi, I just realised I never answered your question, sorry about that. You can find my answer to a similar question [here](https://github.com/MarekKowalski/DeepAlignmentNetwork/issues/52#issuecomment-486982512). Marek

Hi, This whole piece of code implements a method described in Appendix D of an old paper available here: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.141.3089&rep=rep1&type=pdf Hope that clears things up. Marek

You should use the DANtraining.py script. Before you do that however you need to "build" the training set by running: TrainingSetPreparation.py You will also need to download the 300-W training...

Hi, The training is ran for as long as the validation error is decreasing "fairly rapidly". I would finish training if a lower error is not achieved after, let's say...