OSS Alternative - Discover Top Open Source Alternatives to Popular Software

AlibabaResearch/AdvancedLiterateMachinery

A research initiative by Alibaba's Tongyi Lab, focusing on developing advanced AI systems capable of reading, thinking, and creating, with an initial emphasis on sophisticated OCR and document understanding technologies.

Core Features

Comprehensive OCR benchmarking for Large Multimodal Models (CC-OCR).

Generalized text reading across diverse forms with a unified architecture (Platypus).

High-quality visual text generation in complex scenes (SceneVTG).

Automated generation of web page visual presentations (WebRPG).

Unified framework for text spotting, KIE, and table recognition (OmniParser).

Detailed Introduction

Advanced Literate Machinery (ALM) is a pioneering research project from Alibaba's Tongyi Lab, dedicated to building highly intelligent systems that can read, think, and create, potentially surpassing human capabilities. Initially, the project concentrates on enhancing machine literacy through advanced OCR and document understanding. It introduces innovative algorithms and benchmarks like CC-OCR for LMM evaluation, Platypus for generalized text reading, and OmniParser for unified text parsing, pushing the boundaries of AI in visual information processing.