Replies: 3 comments 1 reply
-
嗯,看来纯OCR任务变化不大,增加了新的端到端模型和表格模型。 我近期抽时间研究下 |
Beta Was this translation helpful? Give feedback.
0 replies
-
简单记录下:
|
Beta Was this translation helpful? Give feedback.
1 reply
-
经过一些调查,我们的C++引擎组件也许可以较低成本地升级到 PaddleOCR-2.8 核心推理框架,支持部分最新模型(支持冠军方案的识别模型,但不支持表格模型)。 👉 hiroi-sora/PaddleOCR-json#139 (comment) 比起 2.6 的旧内核,新内核在多方面有提升,但主要缺点是体积 +70MB 。最高倍率打包压缩后,发布体积 +10MB 左右。 感觉升级还是利大于弊,近期有时间的话会弄一弄。 |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
https://github.com/PaddlePaddle/PaddleOCR/releases
PaddleOCR v2.8.0隆重引入了 PaddleOCR 算法模型挑战赛的冠军方案,为 OCR 领域树立了新的标杆。
表格识别方面据说有巨大进步?
作者大佬跟进中?🚀
Beta Was this translation helpful? Give feedback.
All reactions