PaddleOCR tagshttps://gitcode.net/paddlepaddle/PaddleOCR/-/tags2021-04-10T01:45:21+08:00https://gitcode.net/paddlepaddle/PaddleOCR/-/tags/v2.6.0v2.6.0<p data-sourcepos="1:1-1:16" dir="auto"><strong>Release Note</strong></p>
<ul data-sourcepos="2:1-6:2" dir="auto">
<li data-sourcepos="2:1-2:229">Release <a href="/paddlepaddle/PaddleOCR/-/blob/release/2.3/./ppstructure">PP-Structurev2</a>,with functions and performance fully upgraded, adapted to Chinese scenes, and new support for <a href="/paddlepaddle/PaddleOCR/-/blob/release/2.3/./ppstructure/recovery">Layout Recovery</a> and <strong>one line command to convert PDF to Word</strong>;</li>
<li data-sourcepos="3:1-3:164">
<a href="/paddlepaddle/PaddleOCR/-/blob/release/2.3/./ppstructure/layout">Layout Analysis</a> optimization: model storage reduced by 95%, while speed increased by 11 times, and the average CPU time-cost is only 41ms;</li>
<li data-sourcepos="4:1-4:172">
<a href="/paddlepaddle/PaddleOCR/-/blob/release/2.3/./ppstructure/table">Table Recognition</a> optimization: 3 optimization strategies are designed, and the model accuracy is improved by 6% under comparable time consumption;</li>
<li data-sourcepos="5:1-6:2">
<a href="/paddlepaddle/PaddleOCR/-/blob/release/2.3/./ppstructure/kie">Key Information Extraction</a> optimization:a visual-independent model structure is designed, the accuracy of semantic entity recognition is increased by 2.8%, and the accuracy of relation extraction is increased by 9.1%.</li>
</ul>2022-09-08T15:13:36+08:00littletomatodonkeydazhiningsibuqu@163.comhttps://gitcode.net/paddlepaddle/PaddleOCR/-/tags/v2.5.0v2.5.0<p dir="auto"><strong>Release Note</strong></p>
<ul dir="auto">
<li>Release <a href="/paddlepaddle/PaddleOCR/-/blob/release/2.3/./doc/doc_en/ppocr_introduction_en.md#pp-ocrv3">PP-OCRv3</a>: With comparable speed, the effect of Chinese scene is further improved by 5% compared with PP-OCRv2, the effect of English scene is improved by 11%, and the average recognition accuracy of 80 language multilingual models is improved by more than 5%.</li>
<li>Release <a href="/paddlepaddle/PaddleOCR/-/blob/release/2.3/./PPOCRLabel">PPOCRLabelv2</a>: Add the annotation function for table recognition task, key information extraction task and irregular text image.</li>
<li>Release interactive e-book <a href="/paddlepaddle/PaddleOCR/-/blob/release/2.3/./doc/doc_en/ocr_book_en.md"><em>"Dive into OCR"</em></a>, covers the cutting-edge theory and code practice of OCR full stack technology.</li>
</ul>2022-05-10T07:05:02+08:00MissPenguinlichenxia1991@163.comhttps://gitcode.net/paddlepaddle/PaddleOCR/-/tags/v2.1.1v2.1.1<p data-sourcepos="1:1-1:16" dir="auto"><strong>Release Note</strong></p>
<ol data-sourcepos="3:1-6:144" dir="auto">
<li data-sourcepos="3:1-3:159">Newly release model pruning and model quantization tools based on PaddleSlim. <a href="https://github.com/PaddlePaddle/PaddleOCR/tree/release/2.1/deploy/slim" rel="nofollow noreferrer noopener" target="_blank">Path</a>
</li>
<li data-sourcepos="4:1-4:141">Newly release mobile deployment tools based on Paddle-Lite. <a href="https://github.com/PaddlePaddle/PaddleOCR/tree/release/2.1/deploy/lite" rel="nofollow noreferrer noopener" target="_blank">Path</a>
</li>
<li data-sourcepos="5:1-5:133">Newly release Android demo of ppocr system. <a href="https://github.com/PaddlePaddle/PaddleOCR/tree/release/2.1/deploy/android_demo" rel="nofollow noreferrer noopener" target="_blank">path</a>
</li>
<li data-sourcepos="6:1-6:144">Newly release service deployment based on Paddle Serving. <a href="https://github.com/PaddlePaddle/PaddleOCR/tree/release/2.1/deploy/pdserving" rel="nofollow noreferrer noopener" target="_blank">path</a>
</li>
</ol>2021-05-27T02:52:46+08:00xiaoting31891223+tink2123@users.noreply.github.comhttps://gitcode.net/paddlepaddle/PaddleOCR/-/tags/v2.1.0v2.1.0<p data-sourcepos="1:1-1:16" dir="auto"><strong>Release Note</strong></p>
<ol data-sourcepos="3:1-5:156" dir="auto">
<li data-sourcepos="3:1-3:255">Newly release end-to-end text recognition algorithm <a href="https://www.aaai.org/AAAI21Papers/AAAI-2885.WangP.pdf" rel="nofollow noreferrer noopener" target="_blank">PGNet</a> which is published in AAAI 2021. Find tutorial <a href="https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.1/doc/doc_en/pgnet_en.md" rel="nofollow noreferrer noopener" target="_blank">here</a>.</li>
<li data-sourcepos="4:1-4:209">Newly release multi language recognition model, support more than 80 languages recognition. Find tutorial <a href="https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.1/doc/doc_en/multi_languages_en.md" rel="nofollow noreferrer noopener" target="_blank">here</a>.</li>
<li data-sourcepos="5:1-5:156">Optimize the performance of <a href="https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.1/doc/doc_en/models_list_en.md#English" rel="nofollow noreferrer noopener" target="_blank">English recognition model</a>.</li>
</ol>2021-04-20T03:05:35+08:00MissPenguinlichenxia1991@163.comhttps://gitcode.net/paddlepaddle/PaddleOCR/-/tags/v2.0.0v2.0.0<p data-sourcepos="1:1-1:16" dir="auto"><strong>Release Note</strong></p>
<p data-sourcepos="3:1-3:83" dir="auto">一、Support dynamic graph programming paradigm, adapted to Paddle 2.0, including:</p>
<ol data-sourcepos="4:1-13:0" dir="auto">
<li data-sourcepos="4:1-4:38">Detection algorithm: DB, EAST, SAST</li>
<li data-sourcepos="5:1-5:60">Recognition algorithm: Rosetta, CRNN, RARE, SRN, STAR-Net</li>
<li data-sourcepos="6:1-9:42">PPOCR Chinese models:
(1) Detection models: mobile, server
(2) Text direction classification models: mobile
(3) Recognition models: mobile, server</li>
<li data-sourcepos="10:1-13:0">Multilingual models:
(1) English: mobile
(2) Japanese, Korean, French, German, etc. 25 languages in total: mobile</li>
</ol>
<p data-sourcepos="14:1-14:114" dir="auto">二、The related works on deployment have been well adapted, including Inference(Python, C++) , whl, and serving</p>
<p data-sourcepos="16:1-16:52" dir="auto">三、Release the annotation and synthesis tools:</p>
<ol data-sourcepos="17:1-18:196" dir="auto">
<li data-sourcepos="17:1-17:146">Release a new data synthesis tool, i.e., Style-Text,easy to synthesize a large number of images which are similar to the target scene image.</li>
<li data-sourcepos="18:1-18:196">Release a new data annotation tool, i.e., PPOCRLabel, which is helpful to improve the labeling efficiency. Moreover, the labeling results can be used in training of the PP-OCR system directly.</li>
</ol>2021-04-20T03:05:35+08:00zhoujunzjwenmu@gmail.comhttps://gitcode.net/paddlepaddle/PaddleOCR/-/tags/v1.1.0v1.1.0<p data-sourcepos="1:1-1:123" dir="auto">3.5M practical ultra lightweight OCR system, support training and deployment among server, mobile, embedded and IoT devices</p>2021-04-10T01:45:21+08:00dyningdyning.2003@163.com