FuseCap: Leveraging Large Language Models for Enriched Fused Image Captions

Keyphrases

• Fused Image 100%
• Image Caption 100%
• Training Set 50%
• Large-scale Dataset 50%
• Data-driven Approach 50%
• Image Content 50%
• Automated Approach 50%
• Image Details 50%
• Object Detector 50%
• Image Captioning 50%
• Visual Detail 50%
• Pre-training Strategy 50%
• Image Description 50%
• Reannotation 50%
• Vision-language Pre-training 50%
• Optical Character Recognizer 50%
• Image-Text Dataset 50%

Computer Science

• Recognizer 100%
• Large Language Model 100%
• Image Content 50%
• Object Detector 50%
• Image Captioning 50%