Global Invoice Model - Accuracy Benchmark

Following are the accuracy benchmark for Smart Vision 2.6:

Field Names

Accuracy 

Document Type

0.873171

Invoice No

0.892857

Invoice Date

0.977679

PO Number

0.636364

Vendor Name

0.668224

Vendor Code

0.617284

Street/Location

0.804762

ZIP/Pin code

0.77451

City

0.780952

State

0.857143

Vendor Email

0.653465

Vendor Phone

0.775401

Vendor fax

0.722628

Customer Code

0.47541

Customer email

0.517241

Customer phone

0.512195

Ship To Name

0.856383

Street/Location

0.896739

ZIP/Pin code

0.827957

City

0.87766

State

0.899441

Bill To Name

0.888889

Street/Location

0.910377

ZIP/Pin code

0.896396

City

0.928571

State

0.845411

Amount Due

0.837838

Sub Total

0.758794

Tax amount

0.761111

Street/Location

0.586207

ZIP/Pin code

0.653061

City

0.537313

State

0.723404

Bank Name

0.742857

Bank Account

0.738562

Payment Terms

0.88172

Due Date

0.772727