molscribe hangs indefinetly with .tif file from USPTO grant red book

[nbosc]

Despite all my effort to filter out any image not representing a molecular structure in my pipeline, time to time it happens that I submit one of those to MolScribe. For some of them, MolScribe does not return a structure, but it does not return anything either. I can clearly see the RAM consumption increasing but even after hours, MolScribe is still trying (or maybe it is stuck...) Maybe a time limit would help?

Example: