KFUPM Logo

FraudX – Receipt Forgery Detection

KFUPM – ICS 619 | Rayan Alsubhi

New analysis Dashboard Queue 5

Receipt forgery detection for adjudicators & auditors

Upload a receipt image and the model highlights regions that have been modified — copy-paste, deletion, hand-redrawn text, or pixel-level edits — with the type of edit and the field affected (Total / Product / Metadata).

92.25%
Patch precision (vs paper 79.41%)
0.97
Patch-level AUC

Upload receipt

Drag a receipt image here, or click to browse. Analysis runs locally; no images leave this machine.

Analyze = ad-hoc one-off review.  Add to queue = file ends up in the adjudicator queue (drop multiple at once).