New DICModel enhances ICT image captioning with multi-modal LLMs

Researchers have developed a novel Domain-specific Image Captioning Model (DICModel) designed for the ICT industry, utilizing a multi-stage progressive training strategy. This approach combines synthesized image-text pairs with expert annotations to enhance the model's understanding of domain-specific visual information. The DICModel, with only 7 billion parameters, demonstrates superior performance compared to larger state-of-the-art models, significantly improving BLEU metrics and accuracy on domain-specific questions. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT This research could improve the extraction of visual information in specialized domains, potentially advancing multimodal AI capabilities.

RANK_REASON This is a research paper detailing a new model and training methodology. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• DICModel
• ICT
• LLM
• Qwen2.5-VL
• Lianying Chao

• paper
• model release