Where to get the "merged_flame_data.npz"

[niazahamd89]

Hello,

Thank you for your excellent work. I would like to request access to the file "merged_flame_data.npz" from the face dataset. I have sent several requests to the dataset owners but am still awaiting a response.

Could you please let me know if there is any other way to obtain the "merged_flame_data.npz" file?

Thank you very much for your help, and again, I appreciate you sharing such great work.

[moonbow721]

Hello,

Thank you for your kind words and for your interest in our work. The “merged_flame_data.npz” file is not a raw data file, but rather a file we generate by combining the data from subsets of the MICA dataset. Once you have the FLAME parameters for each individual subset, you can use our script at lib/data/face_process/process_mica.py to merge them and create this file yourself.

Regarding access to the MICA dataset, you are right that it requires permission. Due to the sensitive nature of human face data and the dataset's licensing agreements, we are unfortunately unable to release or share the data directly, either as raw files or the merged .npz file.

However, I can suggest the best path forward:

Instead of applying for the original data from each subset (e.g., LYHM, FaceWarehouse), which contains raw RGB images, I would recommend contacting the authors of the MICA paper directly. They should be able to provide you with the fitted FLAME parameters for all the subsets they used.

I hope this suggestion helps you get the data you need. Thank you for your understanding!