Statement fluency-based incomplete Han simplex image conjugation method

• Main Authors: YAO LEI, FAN GAOJUAN, HU RENFEN, LIU HAOXIA, CHEN SHANXIONG, ZHANG CHONGSHENG, CHEN YUANYUAN, MO BOFENG
• Format: Patent
• Language: Chinese; English
• Published: 10.01.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The invention discloses a method for conjugating a broken Chinese simple image based on statement smoothness. The method comprises the following steps: A, carrying out image feature annotation on a to-be-conjugated Chinese simple image; b, constructing a training set and training to obtain a BERT language model after tuning training; c, for the specified broken Han simple image A, obtaining a to-be-conjugated comparison group of the Han simple image A through image feature comparison; and D, performing statement smoothness detection on the Chinese simple image A and the Chinese simple images in the to-be-conjugated comparison group of the Chinese simple image A one by one by using the BERT language model after tuning training to obtain N to-be-selected Chinese simple images with the highest language smoothness with the Chinese simple image A, and taking the N to-be-selected Chinese simple images as the N to-be-selected Chinese simple images with the highest degree of conjugation with the Chinese simple image