A relire : Récupérer le texte d'une image en Javascript avec Tesseract.js (OCR)

https://blog.shevarezo.fr/post/2017/02/10/recuperer-texte-image-javascript-tesseractjs-ocr

#javascript #ocr #tesseract @biject @antimatter15

OCR pipeline for ML training (tables, diagrams, math, multilingual)

https://github.com/ses4255/Versatile-OCR-Program

Qwen-2.5-32B is now the best open source OCR model

https://github.com/getomni-ai/benchmark/blob/main/README.md

#HackerNews #Qwen-2.5-32B #OpenSource #OCR #BestModel #AI #Technology

Microsoft Foto di Windows si rinnova: tante funzioni AI in arrivo
#Aggiornamenti #AI #App #Copilot #EsploraFile #Foto #GommaMagica #IntelligenzaArtificiale #JXL #MicrosoftFoto #Notizie #Novità #OCR #Software #TechNews #Tecnologia #Windows10 #Windows11 #WindowsInsider

https://www.ceotech.it/microsoft-foto-di-windows-si-rinnova-tante-funzioni-ai-in-arrivo/

So if you’re using Mastodon on the web, you can press the ALT button and then follow the “Detect text from picture” link.

On Mac/iOS, you can select text on images as if they were text by clicking/tapping and dragging and paste that in (might be more accurate; that’s what I did).

PS. This was meant to be a reply to https://mastodon.social/@fatbrit/114215995914155838 but somehow didn’t get threaded correctly (was using the web client instead of Mona. I somehow manage to do that there sometimes. Has happened before.) :)

#Шиндошс #NeuralNetwork

В #PowerToys добавили ИИшницу для #OCR и и преобразования текста. Радостно! (#НаСамомДелеНет).

An old photo I ran across

Why extracting data from PDFs is still a nightmare for data experts - For years, businesses, governments, and researchers have struggled with a ... - https://arstechnica.com/ai/2025/03/why-extracting-data-from-pdfs-is-still-a-nightmare-for-data-experts/ #opticalcharacterrecognition #computationaljournalism #largelanguagemodels #machinelearning #simonwillison #derekwillis #raykurzweil #mistralocr #chatgpt #chatgtp #mistral #biz⁢ #tech #pdfs #ocr #ai

Does anybody know what happened to rescribe.xyz?

Mistral #OCR is an Optical Character Recognition API that sets a new standard in document understanding. Unlike other models, Mistral OCR comprehends each element of documents—media, text, tables, equations—with unprecedented accuracy and cognition. It takes images and PDFs as input and extracts content in an ordered interleaved text and images.
As a result, Mistral OCR is an ideal model to use in combination with a #RAG system taking multimodal documents (such as slides or complex PDFs) as input. #AI

https://mistral.ai/fr/news/mistral-ocr

Mistral #OCR, some #AI-based text recognition service, is raking in tons of upvotes on HN.

Top comment: "Impressive model" but "high risk for hallucinations".

WTF! Can you create worse stuff than hallucinating OCR? Stop this AI-addled madness.

https://news.ycombinator.com/item?id=43282905

Mistral OCR API を使って PDF からテキストを抽出する
https://qiita.com/ikuro_mori/items/f428bd207f5588ee3305?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

Mistral lancia API OCR multimodale che trasforma PDF in Markdown. La tecnologia riconosce testo e elementi grafici inserendoli in "bounding box". Il Markdown, formato prediletto dagli LLM, facilita l'elaborazione dei contenuti nei sistemi di IA generativa. #Mistral #OCR #AI

https://youtu.be/6lRBm0KnzBI

Mistral OCR — https://mistral.ai/news/mistral-ocr
#HackerNews #MistralOCR #MistralAI #MachineLearning #OCR #Technology #Innovation

Meine Fresse, sind wir heute wieder aktuell:
#PDF #OCR #CLI #Stapelverarbeitung ... als wäre es 2005 ;)

Wobei: Seit Dokumente zunehmend per Handy "gescannt" werden, könnte (nachträgliche) Texterkennung doch recht aktuell sein :)

https://www.tutonaut.de/pdf-texterkennung-stapelweise-fuer-windows-und-linux/

I'm not convinced: #eScriptorium #LLM enhanced scheint sich darin zu erschöpfen, dass man

1. Transformermodelle für das OCR nutzen könnte (extrem ressourcenintensiv)
2. #OCR correction via #LLM (prompt-based)
3. #NER mit #LLM (prompt-based)

Nichts von dem erscheint mir in meiner Naivität den sozialen, ökonomischen und ökologischen Impact der #LLM Nutzung zu rechtfertigen. Und ich werd auch nicht warm mit diesem Prompting-Ansatz.