visual-rag
retail_visual_rag_pipeline
A Visual RAG Pipeline for Few-Shot Fine-Grained Product Classification
Paper
Accepted at The 12th Workshop on Fine-Grained Visual Categorization (FGVC12) at IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2025.
Overview
The Retail Visual RAG Pipeline Dataset is a subset of the Retail-786k (Retail-786k) image dataset, supplemented with additional textual data per image.
Data
Image Data:
The images are cropped from scanned… See the full description on the dataset page: https://huggingface.co/datasets/blamm/retail_visual_rag_pipeline.Visual-RAG-MEOfficial data for Visual-RAG-ME, a multi-entity text-to-image retrieval and VQA dataset. Please check our paper for more details.
visual-rag-data