Upload a tablet photo. The local/online backend will receive the image and return a proposed ADS structure.
Use this field later to correct AI output. Corrections will become training data.