PAIRS DATA AND CODE

PAIRS: Panofskian Art Interpretation and Retrieval System Repository Contents This repository contains the complete implementation, datasets, and experimental data for the paper "PAIRS: A Panofskian Art Interpretation and Retrieval System Based on Multi-Agent Collaboration". File Structure File Size Description code.zip - Complete source code for PAIRS annotation-data.zip - Pre-computed annotations for all artworks semart-images-1.zip - SemArt dataset images (Part 1) semart-images-2.zip - SemArt dataset images (Part 2) semart-files.zip - SemArt dataset metadata and query files SUPPLEMENTARY_MATERIALS.md - Supplementary materials for the paper 1. Code (code.zip) Complete implementation of the PAIRS framework. Structure code/ ├── art-retrieval-research/ │ └── data_annotation_system/ # MC-ArtAnnotate algorithm implementation ├── image_retrieval_platform/ # FastAPI-based retrieval platform └── experiment/ # Experiment scripts and baselines Components Data Annotation System: Multi-agent collaborative annotation framework with six specialized agents (Form Analysis, Subject Identification, Symbolism, Cultural Context, Technical Material, Coordinator). Image Retrieval Platform: REST API for art image retrieval with LLM-based verification. Experiment Code: Scripts for reproducing all experimental results, including baseline implementations (GPT-4o, Claude Sonnet 4.5, InternVL-2.5, BLIP-2, CLIP-Large). See individual subdirectory README files for detailed usage instructions. 2. Annotation Data (annotation-data.zip) Pre-computed multi-agent annotations for all artworks in the evaluation datasets. Contents Structured annotations following Panofsky's three-level iconographic framework Global annotations (form, subject, symbolism, cultural context, technical analysis) Regional annotations with spatial coordinates Vector embeddings (BGE-large-en-v1.5, 1024 dimensions) 3. SemArt Dataset The SemArt benchmark is split into three files due to size constraints. Images semart-images-1.zip: Artwork images (Part 1) semart-images-2.zip: Artwork images (Part 2) Metadata (semart-files.zip) Image metadata (title, artist, date, type) Official train/test splits 100 evaluation queries with ground-truth annotations Dataset Scales We construct three evaluation scales from SemArt: Dataset Images Purpose data-100 100 Rapid validation data-1069 1,069 Primary benchmark (complete test set) data-5000 5,000 Scalability testing License SemArt is released under CC BY-NC 4.0 (Creative Commons Attribution-NonCommercial). See the original dataset page for details. All experimental datasets are derived from the publicly available SemArt benchmark. The 100 evaluation queries are drawn from SemArt's official test partition, testing iconographic understanding, symbolic interpretation, and cultural context comprehension. Image Sources: Artwork images in SemArt originate from museum digital collections including the Web Gallery of Art and Wikimedia Commons. Many source images are in the public domain due to artwork age (pre-1900). 4. PAIRS Output and Copyright Annotation Outputs PAIRS generates original textual annotations through LLM-based analysis. These annotations constitute derivative scholarly analysis rather than reproduction of copyrighted content. Annotations describe visual elements, interpret symbolic meanings, and provide cultural context—all original analytical text generated by our system. No Image Reproduction Our system does not reproduce, modify, or redistribute copyrighted artwork images. During retrieval, the system returns references to original images (URLs) rather than copies. Users access images through their original sources. Model Outputs LLM-generated content (annotations and verification reasoning) represents original analytical work. These outputs do not reproduce protected expression from training data but synthesize domain knowledge into new scholarly descriptions. System Requirements Python: 3.10+ GPU: CUDA-compatible GPU (recommended for embedding computation) RAM: Minimum 16GB for data-1069 dataset Minimum 64GB for data-5000 dataset API Keys: Anthropic, OpenAI, or compatible LLM services Additional Documentation SUPPLEMENTARY_MATERIALS.md - Detailed supplementary materials including dataset licensing, implementation notes, and best practices