FAU
FAU copied to clipboard
EvelynFan
Json file
Hello and thank you very much for sharing this project. I have a DISFA dataset, that means I have Landmark positions and UA intensities. I have also the videos and the images of DISFA dataset. But I don't have any Json file. You noted that we should make them by hand, but for 27 candidates which every one has 4845 frames and 66 landmarks as well as 27 AU files with 12 AUs for each, how do I find the coordinates of the AUs?
Can you guide me please?
Regards and thanks again
Hello I am looking forward to hearing from you.
Regards
Hi, you can use an automatic facial landmark detector to locate the coordinates of the facial landmarks for each frame. I defined the coordinates of the AUs based on where the action unit occurs. FYI, you may look at Ntinou et al.'s work (https://arxiv.org/pdf/2004.06657.pdf) for more details. Table 1 shows the correspondences between landmarks and AUs.
Perfect! Thank you very much for your answer. In fact your paper and code will be an important part of my job and I would be happy if you give me more advice on how to use it. @EvelynFan
I already use OpenFace which gives me the Landmarks coordinates. I will therefore use the table1 information and prepare the Json file.
About the Json file, please note that in the address of each image we should not give a full address of them, because in "function.py" there is a variable named "path_new" which starts from the current address of "function.py" so if we give a full address in the Json file, we will get the "could not read" error!