Table 6 The results of the text description generator.

From: A customized image editing framework for diverse prohibited and restricted products in illegal online transactions

Model

Recognition accuracy (%)

BLEU (%)

MobileVLM

7.50

9.75

MobileVLM-Fine-tuned

16.69

16.18

CLIP + MobileVLM-Fine-tuned

78.38

36.65

Back to article page