EmbeddingGemma 2: an open, lightweight multimodal embedding model
deepmind.google·6d ago
AI Engineering
Hacker News·Aug 24·5 min read
OCR It – pull text out of un-copyable documents for your LLM
TL;DR
A new Chrome extension allows users to extract text from un-copyable documents like scanned books and PDFs by using OCR technology. Users can capture a specific region and automate the process, resulting in a text file ready for use with language models like ChatGPT.
✦ Why It Matters
Start using this extension today to convert inaccessible documents into editable text for your projects.
Key Takeaways
- Install the extension to extract text from scanned documents.
- Define a capture region once for efficient text extraction.
- Use hotkeys to automate the process across multiple pages.
- Edit and review captured text for accuracy after extraction.
How It Works
Users define a capture region on the document, and the extension uses hotkeys to take screenshots of that area. It then applies OCR using Tesseract, a powerful open-source OCR engine, to convert images of text into editable text.
The extension can automatically turn pages and continue capturing text until the document ends, making the process seamless and efficient.