# dwzhu-pku/paperbanana **Attribution required: if you use, quote, or summarise this content, you must credit and link back to [awesome-repositories.com](https://awesome-repositories.com/repository/dwzhu-pku-paperbanana).** 3,742 stars · 174 forks · JavaScript ## Links - GitHub: https://github.com/dwzhu-pku/PaperBanana - Homepage: https://dwzhu-pku.github.io/PaperBanana/ - awesome-repositories: https://awesome-repositories.com/repository/dwzhu-pku-paperbanana.md ## Description PaperBanana is an AI research visualization tool and framework designed to generate and refine high-resolution academic illustrations from conceptual and technical descriptions. It employs an automated generation pipeline that transforms scientific text and captions into publication-quality diagrams and plots. The system utilizes a multi-stage process consisting of retrieval-augmented planning, image synthesis, and a critic-based iterative refinement mechanism. This workflow allows for the adjustment of image details and the upscaling of visual outputs to 4K resolution. The project includes capabilities for batch-parallel image synthesis to facilitate comparative selection and prototyping. It further provides a scientific image evaluation framework to measure output quality against ground-truth datasets and tracks the evolution of images through intermediate state visualization. ## Tags ### Artificial Intelligence & ML - [Academic Visuals](https://awesome-repositories.com/f/artificial-intelligence-ml/academic-paper-generators/academic-visuals.md) — Generates publication-quality academic architecture diagrams and data-driven charts from scientific text. ([source](https://cdn.jsdelivr.net/gh/dwzhu-pku/paperbanana@main/README.md)) - [Iterative Refinement Workflows](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/agentic-workflows/iterative-refinement-workflows.md) — Employs a critic-based feedback loop to iteratively refine visual outputs and correct errors. - [AI Research Tools](https://awesome-repositories.com/f/artificial-intelligence-ml/ai-research-tools.md) — A specialized framework for AI-driven automated design and generation of high-resolution research visualizations. - [Quality Evaluators](https://awesome-repositories.com/f/artificial-intelligence-ml/generative-content-apis/quality-evaluators.md) — Provides a measurement framework to evaluate the prompt alignment and aesthetic quality of generated academic diagrams. ([source](https://cdn.jsdelivr.net/gh/dwzhu-pku/paperbanana@main/README.md)) - [Technical Diagram Refinement](https://awesome-repositories.com/f/artificial-intelligence-ml/generative-upscalers/sequential-detail-refinement/technical-diagram-refinement.md) — Improves the detail and resolution of generated technical illustrations through iterative editing and upscaling. - [Visual Detail Refinement](https://awesome-repositories.com/f/artificial-intelligence-ml/generative-upscalers/sequential-detail-refinement/visual-detail-refinement.md) — Refines visual details and upscales resolution to 4K using a critic mechanism and text-based editing. ([source](https://cdn.jsdelivr.net/gh/dwzhu-pku/paperbanana@main/README.md)) - [Layout Planning Augmentation](https://awesome-repositories.com/f/artificial-intelligence-ml/language-model-orchestration/retrieval-augmented-generation/layout-planning-augmentation.md) — Uses external data sources to inform the layout and conceptual structure of diagrams before rendering. - [Illustration Quality Evaluators](https://awesome-repositories.com/f/artificial-intelligence-ml/machine-learning/infrastructure/model-evaluation-and-validation/model-benchmarking/scientific-model-evaluators/illustration-quality-evaluators.md) — Provides a set of metrics and tools for measuring the quality of AI-generated academic illustrations against ground-truth datasets. - [Scientific Asset Metrics](https://awesome-repositories.com/f/artificial-intelligence-ml/model-performance-evaluators/generative-asset-quality-metrics/scientific-asset-metrics.md) — Includes a framework to measure the accuracy of generated illustrations against ground-truth datasets using performance indicators. - [Generation Evolution Tracking](https://awesome-repositories.com/f/artificial-intelligence-ml/generation-evolution-tracking.md) — Tracks the intermediate pipeline steps showing how a conceptual description evolves into a final academic illustration. ([source](https://cdn.jsdelivr.net/gh/dwzhu-pku/paperbanana@main/README.md)) - [Batch Processing](https://awesome-repositories.com/f/artificial-intelligence-ml/image-translation-pipelines/image-text-translators/batch-processing.md) — Provides automated execution of illustration pipelines across multiple candidates via batch processing. ([source](https://cdn.jsdelivr.net/gh/dwzhu-pku/paperbanana@main/README.md)) ### Data & Databases - [Visual Generation Pipelines](https://awesome-repositories.com/f/data-databases/data-processing-pipelines/document-llm-preparation/multi-stage-pipeline-processing/visual-generation-pipelines.md) — Transforms scientific text into images through a sequential multi-stage pipeline of retrieval, planning, and visualization. ### Graphics & Multimedia - [Academic Figure Pipelines](https://awesome-repositories.com/f/graphics-multimedia/graphics-and-media/vector-graphics-resources/academic-figure-pipelines.md) — Implements a pipeline that converts academic captions into publication-quality figures via sequential synthesis. - [Visual Research Prototyping](https://awesome-repositories.com/f/graphics-multimedia/visual-research-prototyping.md) — Generates multiple candidate illustrations simultaneously to compare and select the most effective visual for a study. ### Scientific & Mathematical Computing - [Scientific Visualizations](https://awesome-repositories.com/f/scientific-mathematical-computing/scientific-visualizations.md) — Transforms complex research data and conceptual descriptions into clear scientific visual representations for papers. ### Part of an Awesome List - [Batch Parallel Synthesis](https://awesome-repositories.com/f/awesome-lists/ai/image-generation-and-synthesis/batch-parallel-synthesis.md) — Implements batch-parallel synthesis to allow users to compare multiple candidate illustrations for a single concept. ### User Interface & Experience - [Generation Step Visualizers](https://awesome-repositories.com/f/user-interface-experience/real-time-visual-inspectors/state-transition-visualizers/generation-step-visualizers.md) — Provides intermediate state visualization to track how a conceptual description evolves into a final academic illustration.