Fine-tuning InternVL2 on generated interleaved image-text conversations substantially improves contextual image referencing in RAG chatbots, with a new benchmark and metrics.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
ImageRef-VL: Enabling Contextual Image Referencing in Vision-Language Models
Fine-tuning InternVL2 on generated interleaved image-text conversations substantially improves contextual image referencing in RAG chatbots, with a new benchmark and metrics.