I have been invited to give a keynote (my first one!) at the Transkribus User Conference this September. I wanted to attend this conference for a long time, so being able to join in this role is a special honor to me. Hope to see many of you in Passau at the TUC! #digitalhumanities #dh #multilingualism #multilingualdh #textrecognition #transkribus
Link: https://tuc.transkribus.org/programme/keynotes/keynote-aliz
Full program: https://tuc.transkribus.org/programme/

FOSS Advent Calendar - Door 11: Read Any Text with EasyOCR

Meet EasyOCR, a lightweight open source optical character recognition (OCR) engine that makes extracting text from images and documents almost effortless. Supporting over 80 languages, including those with complex scripts and mixed language text, it's designed to be powerful, accurate, and incredibly straightforward to use.

Built on PyTorch and integrating deep learning models, EasyOCR delivers high recognition accuracy even on challenging images, low resolution, skewed text, or complex backgrounds. What sets it apart is its simplicity: with just a few lines of code, you can have a fully functional OCR pipeline running locally, without needing an internet connection or external APIs. Your data remains completely private.

Whether you're digitizing printed material, extracting text from screenshots (for example, lyrics from L’âme Immortelle, an Austrian dark wave band), automating document workflows, or analyzing visual data, EasyOCR gets the job done quickly and reliably.

Pro tip: Use it to create searchable PDFs, translate foreign text in images, or even capture and digitize handwritten notes with the right training data.

Link: https://github.com/JaidedAI/EasyOCR

What text would you like to extract from images? Scanned books, street signs, or maybe your old family documents?

#FOSS #OpenSource #OCR #EasyOCR #TextRecognition #AI #DeepLearning #Python #ComputerVision #DocumentDigitization #DataExtraction #Privacy #LocalAI #Multilingual #OpenTools #Fediverse #TechNerds #AdventCalendar #adventkalender #adventskalender #TextExtraktion #KI #PyTorch #DevCommunity #Automation #OfflineAI #PythonProgramming

🤖: "Why not just treat text as an image?" - A thought-provoking idea, if only this groundbreaking revelation hadn't already been around since the dawn of #OCR. 📚 Because clearly, the future of #AI hinges on converting text into... more text, but slower. 🐢🚀
https://www.seangoedecke.com/text-tokens-as-image-tokens/ #Innovation #TextRecognition #FutureTech #HackerNews #ngated

Coming up on October 3-4, 2025 at Central European University in Vienna: OCR/HTR Workshop for Under-resourced and Under-represented Languages in Digital Humanities, funded by the Cluster of Excellence EurAsian Transformations and by CLARIAH-AT! (Main organizer: yours truly) #digitalhumanities #multilingualdh #textrecognition #ocr #htr

Wow, who knew that plain old text recognition could be such an ✨adventure✨? Apparently, adding a sprinkle of "accuracy" to your #OCR code is groundbreaking 🤯. But hey, the world couldn't possibly survive without yet another way to read printed text, right? 📚🔍
https://pieces.app/blog/how-we-made-our-optical-character-recognition-ocr-code-more-accurate #textrecognition #adventure #technology #innovation #accuracy #printedtext #HackerNews #ngated