{"id":3834,"date":"2023-10-31T11:00:00","date_gmt":"2023-10-31T02:00:00","guid":{"rendered":"http:\/\/52.198.218.71\/?post_type=news&p=3834"},"modified":"2023-11-13T13:36:47","modified_gmt":"2023-11-13T04:36:47","slug":"flax-scanner-for-sd","status":"publish","type":"news","link":"https:\/\/cinnamon.ai\/en\/news\/flax-scanner-for-sd\/","title":{"rendered":"\u201cFlax Scanner for SD Invoice General-purpose Model\u201d will be on sale from November 6th"},"content":{"rendered":"

Invoice general-purpose AI-OCR compatible with the Electric Book Act and the Invoice System
\u201cFlax Scanner for SD Invoice General-purpose Model\u201d will be on sale from November 6th
Reading accuracy 93%*1<\/sup>\/<\/strong> Automatically reads 27 items from invoices in different formats<\/strong><\/p>\n\n\n\n

<\/p>\n\n\n

\n
\"\"<\/figure>\n<\/div>\n\n\n

\u3000<\/p>\n\n\n\n

Our company has developed an original AI-OCR "Flax Scanner for SD Invoice General Model" that automatically recognizes 27 items and extracts text from atypical invoices, which have different formats for each business partner, without having to define coordinates in advance. On sale from November 6, 2023.

The development of an AI-OCR system that can read documents with countless formats, such as invoices, requires advanced AI-OCR technology (feature learning type*) that does not require the coordinate definition of extraction items.2<\/sup>) is required. The feature-learning AI-OCR "Flax Scanner for SD Invoice General-purpose Model" announced this time can automatically read 27 items commonly used in invoices (*see attached document). Accuracy 93%*1<\/sup>and achieves high-precision reading. Cinnamon AI is a highly versatile and highly accurate AI-OCR that can be quickly introduced without the cost and time required to build a dedicated model, and is promoting significant efficiency improvements in invoice processing operations. Masu.<\/p>\n\n\n\n

*1 All values are Cinnamon AI test data values.
*2 Feature learning AI-OCR is a technology that can identify items by learning the various common characteristics of each item in multiple forms, and can link written content and items. Since it automatically determines what type of document it is and what is written where, there is no need for any configuration by the customer. Therefore, it can be said that AI-OCR is strong in handling non-standard forms in a wide variety of formats.<\/p>\n\n\n\n

<\/p>\n\n\n\n

\u25a0Development background<\/strong><\/p>\n\n\n\n

With the introduction of the Electric Book Act and the invoice system, the digitization and electronic transmission of invoices has become increasingly important. Companies are required to store applicable documents as electronic ledgers and maintain data accuracy. On the other hand, in the accounting field, a large number of invoices are received every month in various formats, and it takes a great deal of time and effort to check the contents and input data manually, increasing the operational burden and slowing down the invoice processing operations. The need for efficiency and accurate data processing is increasing.<\/p>\n\n\n\n

Conventionally, in order to convert invoices into data using AI-OCR, it has been common to use a coordinate definition type in which reading points and items are set in advance for each format, but it takes a lot of man-hours to set up in advance, and it also slows down transactions. The challenge was that we had to create countless templates because each destination had a different format.<\/p>\n\n\n\n

Until now, our company has been able to automatically read non-standard forms including invoices by developing a customer-specific AI-OCR model and providing it to various companies. Development was expensive, and the number of companies that could implement it was limited.<\/p>\n\n\n\n

In order to introduce AI-OCR to more companies, we have started to develop a "general-purpose invoice model" by leveraging the know-how we have cultivated so far, and we are using Cinnamon AI's advanced AI technology to support various invoice formats. A practical-level general-purpose AI-OCR that can read data has been completed and is now being announced.<\/p>\n\n\n\n

<\/p>\n\n\n\n

\u25a0Characteristics of general-purpose invoice model<\/strong><\/p>\n\n\n\n

- No need to create templates
<\/strong>Regardless of the format, AI determines what is written where on the invoice, and extracts and reads each item with high precision. In addition to the business partner, registration number, and transaction date,
It is possible to read up to 27 items such as product name, quantity, and unit price included in the details, and the reading results can be output as a CSV file or linked via API.<\/p>\n\n\n\n

\u30fbCan be read with high accuracy
<\/strong>It achieves highly accurate reading of up to 27 items with an average character accuracy of 92.99%* and an average item-specific accuracy of 86.17%*.
*Both are Cinnamon AI test data values<\/p>\n\n\n\n

\u30fb Form classification function
<\/strong>Since uploaded invoices and other forms can be automatically classified, it is also possible to handle forms other than invoices. After classification, AI can automatically determine and process the optimal AI-OCR model provided by Cinnamon AI (general-purpose invoice model, coordinate-defined model, individually customized model, etc.).<\/p>\n\n\n\n

\u25c7 For inquiries from companies regarding this matter: https:\/\/go.cinnamon.ai\/inquiry.html<\/a><\/p>\n\n\n\n

<\/p>\n\n\n\n

<\/p>\n\n\n\n

<Reference materials>
\u00a027 items to be read by \u201cFlax Scanner for SD Invoice General Model\u201d<\/strong><\/p>\n\n\n\n

\u25c7Reading accuracy (test data value by Cinnamon AI)
<\/strong>Accuracy per character: Average 92.99% \/ Accuracy by item: Average 86.17%<\/p>\n\n\n\n

No.<\/strong><\/td>Reading item name<\/strong><\/td><\/tr>
1<\/td>Billing company name<\/td><\/tr>
2<\/td>Billing organization name<\/td><\/tr>
3<\/td>Billing phone number<\/td><\/tr>
4<\/td>Billing fax number<\/td><\/tr>
5<\/td>Billing postal code<\/td><\/tr>
6<\/td>Billing address<\/td><\/tr>
7<\/td>Registration number<\/td><\/tr>
8<\/td>date of issue<\/td><\/tr>
9<\/td>Delivery date<\/td><\/tr>
10<\/td>invoice number<\/td><\/tr>
11<\/td>document number<\/td><\/tr>
12<\/td>Date of payment<\/td><\/tr>
13<\/td>Billing company name<\/td><\/tr>
14<\/td>Billing organization name<\/td><\/tr>
15<\/td>Total amount excluding tax<\/td><\/tr>
16<\/td>consumption tax<\/td><\/tr>
17<\/td>Total amount including tax<\/td><\/tr>
18<\/td>Bank name*<\/td><\/tr>
19<\/td>Branch name*<\/td><\/tr>
20<\/td>Account type*<\/td><\/tr>
21<\/td>account number*<\/td><\/tr>
22<\/td>Account name*<\/td><\/tr>
23<\/td>Product name (details)<\/td><\/tr>
24<\/td>Quantity (details)<\/td><\/tr>
25<\/td>Unit (details)<\/td><\/tr>
26<\/td>Unit price (details)<\/td><\/tr>
27<\/td>Amount (details)<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

* Up to 5 readings are possible for each item.
Items marked with (details) are read in table format.<\/p>","protected":false},"excerpt":{"rendered":"

\u96fb\u5e33\u6cd5\u3068\u30a4\u30f3\u30dc\u30a4\u30b9\u5236\u5ea6\u306b\u5bfe\u5fdc\u3057\u305f\u8acb\u6c42\u66f8\u6c4e\u7528AI-OCR\u300cFlax Scanner for SD \u8acb\u6c42\u66f8\u6c4e\u7528\u30e2\u30c7\u30eb\u300d\u309211\u67086\u65e5\u3088\u308a\u8ca9\u58f2\u8aad\u307f\u53d6\u308a\u7cbe\u5ea693\uff05*1\/ \u30d5\u30a9\u30fc\u30de\u30c3\u30c8\u306e\u7570\u306a\u308b\u8acb\u6c42\u66f8\u306e27\u9805\u76ee\u3092\u81ea\u52d5\u8aad\u307f\u53d6\u308a […]<\/p>","protected":false},"template":"","news-cat":[27],"news-tag":[],"class_list":["post-3834","news","type-news","status-publish","hentry","news-cat-press"],"acf":[],"aioseo_notices":[],"_links":{"self":[{"href":"https:\/\/cinnamon.ai\/en\/wp-json\/wp\/v2\/news\/3834","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/cinnamon.ai\/en\/wp-json\/wp\/v2\/news"}],"about":[{"href":"https:\/\/cinnamon.ai\/en\/wp-json\/wp\/v2\/types\/news"}],"wp:attachment":[{"href":"https:\/\/cinnamon.ai\/en\/wp-json\/wp\/v2\/media?parent=3834"}],"wp:term":[{"taxonomy":"news-cat","embeddable":true,"href":"https:\/\/cinnamon.ai\/en\/wp-json\/wp\/v2\/news-cat?post=3834"},{"taxonomy":"news-tag","embeddable":true,"href":"https:\/\/cinnamon.ai\/en\/wp-json\/wp\/v2\/news-tag?post=3834"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}