PaddleOCR
PulseAugur coverage of PaddleOCR — every cluster mentioning PaddleOCR across labs, papers, and developer communities, ranked by signal.
5 day(s) with sentiment data
-
llama.cpp releases new versions, openai-python updates API features
The llama.cpp project has released two new pre-release versions, b11972 and b11946. Version b11972 includes updates to common and mtmd components, addressing issues with control vector layer indices and M-RoPE position …
-
Tencent unveils WeVisOne-VL, unifying vision tasks into single token stream
Tencent's AI lab has unveiled WeVisOne-VL, a novel vision-language model capable of simultaneously predicting bounding boxes, masks, and text from a single input stream. This unified approach, demonstrated on an RTX 409…
-
AI workflow extracts food package data for South African warning criteria
Researchers have developed a new workflow for extracting evidence from food-package photographs to comply with South Africa's draft R3337 warning criteria. This system uses a region-guided detector and OCR to identify a…
-
AI workflow aids food-package evidence extraction for South African regulations
Researchers have developed a workflow to extract evidence from food-package photographs to comply with South Africa's draft R3337 warning criteria. The system first identifies and extracts information related to nutriti…
-
AI model auto-approves 86.8% of synthetic Japanese invoices
A vision-language model fine-tuned with QLoRA has demonstrated an 86.8% auto-approval rate on synthetic Japanese qualified invoices. The system uses a combination of schema validation, check digits, registry lookups, an…
-
Social Sciences DataLab releases fast Indra OCR model
Social Sciences DataLab has released Indra OCR, a new 1-million-parameter OCR model designed for high throughput and speed on CPUs. It can process up to 49 pages per second, significantly outperforming Tesseract and Pad…
-
OCR and LLM pipeline improves contract data extraction
A developer has created a pipeline to improve the extraction of data from complex legal and financial documents. This system combines Layout Aware OCR tools like Docling and PaddleOCR with Large Language Models such as …
-
LLMs struggle with noisy documents, new benchmark reveals
A new research paper benchmarks several open-source large language models (LLMs) for key-value pair extraction from documents, specifically examining their performance under Optical Character Recognition (OCR) noise. Th…
-
AI Blog Posts Cover OCR, Enterprise AI, and GPU Management · 3 sources tracked
This cluster highlights several AI-related blog posts shared on Mastodon. One post discusses the release of PaddleOCR 3.5, which utilizes a Transformers backend for OCR and document analysis tasks. Another article explo…
-
MLLMs struggle with low-resource Khmer documents, study finds
A new pilot study has evaluated the capabilities of multimodal large language models (MLLMs) in understanding low-resource Khmer documents. Researchers found that while current MLLMs can process visually clear English a…
-
PaddleOCR 3.5 integrates transformers backend; GPU management discussed
PaddleOCR has released version 3.5, now featuring a transformers backend for enhanced OCR and document analysis capabilities. This update leverages advanced transformer architectures to improve performance on these task…
-
Hugging Face blog posts cover AI openness, model updates, and routing · 6 sources tracked
Hugging Face is publishing a series of blog posts covering various AI topics. These include discussions on the importance of openness in AI and cybersecurity, the addition of MCP tools to Reachy Mini, and an analysis of…
-
Hugging Face highlights new AI tools: PaddleOCR, Holo3.1, and VoiceEQ
Hugging Face is highlighting several new AI tools and research. PaddleOCR 3.5 has been released with a Transformers backend for OCR and document analysis. Holo3.1 is presented as a fast, local computer utilization agent…
-
PaddleOCR 3.5 integrates Transformers backend for enhanced OCR
PaddleOCR has released version 3.5, which now utilizes a Transformers backend. This update enhances its capabilities for OCR and document analysis tasks. The new version is available via Hugging Face.
-
RAG tools automate pipeline selection but lag on OCR capabilities
Three open-source tools—AutoRAG, RAGBuilder, and Red Hat AutoRAG—aim to simplify the process of building effective Retrieval-Augmented Generation (RAG) pipelines by automating the testing and selection of optimal config…
-
Hugging Face features PaddleOCR 3.5, JetBrains' Mellum2, and IBM's ScarfBench
Hugging Face is highlighting several new AI developments. PaddleOCR has released version 3.5, incorporating a Transformers backend for enhanced OCR and document analysis. JetBrains has introduced Mellum2, a large expert…
-
Baidu releases Unlimited OCR, challenging long-context AI memory mechanisms · 1 source tracked
Baidu has open-sourced a new OCR model called Unlimited OCR, which excels at processing long documents by mimicking human reading habits. Unlike traditional OCR systems that process documents page by page and then stitc…
-
University seeks on-premise document parsing tools for data governance
A university IT department is seeking an on-premise document processing solution to index and search administrative PDFs, class schedules, and meeting notes. Due to data governance policies, cloud-based APIs are not an …
-
PaddleOCR releases PP-OCRv6 with 50-language support on Hugging Face
PaddleOCR has released PP-OCRv6, an updated suite of universal OCR models available on Hugging Face. This new generation offers improved text detection and recognition accuracy, with models ranging from 1.5 million to 3…
-
JetBrains unveils Mellum2 expert model; PaddleOCR 3.5 adds Transformer backend
JetBrains has launched Mellum2, a 120 billion parameter Mixture-of-Experts model. Separately, PaddleOCR has released version 3.5, which utilizes a Transformers backend for OCR and document analysis tasks. Both announcem…