Supplier qualification auditing method and auditing device based on double-model fusion

The invention provides a supplier qualification auditing method based on double-model fusion, which comprises the following steps: establishing a data set D1 for a target detection algorithm Light-BidDet and a data set D2 for a Chinese and English character recognition algorithm E2E-BidOcr, and divi...

Ausführliche Beschreibung

Gespeichert in:

Bibliographische Detailangaben
Hauptverfasser:	LIN RUIAN, WU GAOJIE, NA KANG-YUN, NI SHAOFENG, LIANG HAITAO, GUO SUQIN, CHEN KUN
Format:	Patent
Sprache:	chi ; eng
Schlagworte:	CALCULATING COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS COMPUTING COUNTING DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES ELECTRIC DIGITAL DATA PROCESSING HANDLING RECORD CARRIERS PHYSICS PRESENTATION OF DATA RECOGNITION OF DATA RECORD CARRIERS SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR
Online-Zugang:	Volltext bestellen
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Beschreibung
Zusammenfassung:	The invention provides a supplier qualification auditing method based on double-model fusion, which comprises the following steps: establishing a data set D1 for a target detection algorithm Light-BidDet and a data set D2 for a Chinese and English character recognition algorithm E2E-BidOcr, and dividing the two data sets into a training set and a verification set according to a proportion of 5: 1; performing supervised training on the Light-BidDet model and the E2E-BidOcr model on the training set of the data set D1 and the data set D2, and performing verification on the verification set to obtain the trained Light-BidDet model and the E2E-BidOcr model; the trained target detection algorithm Light-BidDet model is used for carrying out positioning detection on a designated area of the bidding document file, then the trained end-to-end Chinese and English character recognition algorithm E2E-BidOcr is used for recognizing characters in the area needing to be recognized in the bidding document file, and a recogni