Show HN: Image Categorizer – sort a photo library with local vision models (github.com)

A new tool, Image Categorizer, has been released that leverages local AI models to efficiently organize photo libraries by categorizing images based on their content. Utilizing the Ollama server, the application employs a vision model to analyze each photo, flagging accidental shots, and a text model that sorts images into user-defined categories. After processing, users can review the results through an interactive browser report, where they can make adjustments before exporting a script that moves photos into categorized folders and updates their captions without moving or deleting any original files. This development is significant for the AI/ML community as it simplifies the management of large photo libraries, particularly addressing the challenge of sorting through a mix of valuable and accidental shots. By using local models, it ensures user privacy since no photo leaves the local network. With the ability to customize and save categorization rules, the tool streamlines ongoing organization and supports quick re-sorting without needing to reanalyze images, making it a valuable resource for users with extensive photo collections. Additionally, its efficient processing times, as evidenced by sorting over a thousand photos in under an hour, highlights the practical implications of integrating AI technologies in everyday tasks.