百度开源 Unlimited OCR 2 长文档识别模型

百度发布并开源 Unlimited OCR 2，采用 SR-SWA 机制把文本-视觉注意力交互压缩为固定大小的工作引用，在有限上下文窗口内解析几乎任意长度的文档。随着文档变长，每步解码延迟和内存占用接近恒定，模型在 MPDocBench 和 MarathonDocBench 上取得 SOTA 表现，覆盖几十到数百页文档。

另外 2 家也报道了正文 · 原文Unlimited OCR 2 is here, and it's open source. 📖 ✨

Longer documents can mean a lot more attention computation. Our new SR-SWA mechanism keeps that in check by condensing text-visual attention interactions into a fixed-size set of working references.

Here's what it delivers:- Document parsing at virtually any length, within a finite context window- Nearly constant per-step decoding latency and memory footprint as documents grow longer- State-of-the-art performance on MPDocBench and MarathonDocBench, covering documents from dozens to hundreds of pages

Check out the model here ↓

来源：Baidu Inc. · x.com