user1018

Results 84 comments of user1018

does StyleText support long inputs with spaces between words?

Spaces are not supported in styleText. For space recognition, , the blank area around a single word is relatively large, and may not be good effect. You can add a...

Check failed: (kernels.size() > 0): 0!>0 Error: This model is not supported, because kernel for 'io_copy' is not supported by Paddle-Lite.

预测库和Paddlelite版本一致吗

更换backBone

换backbone之前有在CRNN训练吗，确保数据和标注文件没有问题

enhanced_ctc_loss

现在ch_PP-OCRv2_rec_enhanced_ctc_loss.yml使用的就是enhanced_ctc_loss，可以参考[enhanced_ctc_loss文档](https://github.com/PaddlePaddle/PaddleOCR/blob/release%2F2.6/doc/doc_ch/enhanced_ctc_loss.md)，其中C-CTC是CTC Loss + Center Loss，效果最好，所以采用C-CTC

Which external software is used to edit Excel for table annotation? The Excel cannot pop up when I clicked on the Table Recognition button in the upper right corner of PPOCRLabel

Using the table recognition model in [PP-Structure](https://github.com/PaddlePaddle/PaddleOCR/blob/release%2F2.6/ppstructure/table/README.md). The platform you’re using is Windows, Linux or Mac? For more information, please refer to [PPOCRLabel document](https://github.com/PaddlePaddle/PaddleOCR/blob/release%2F2.6/PPOCRLabel/README.md)

How to use StyleText to generate text image with your own font？

Training code is not provided at this stage

怎么设置学习率下降的速度

就是学习率初始值不变，怎么调整参数加速训练吗

使用自己造的数据集训练与测试集准确率较高，再用另一种情境下的扫描笔真实数据测试准确率几乎为0。

1）推理的时候，字典更换了吗 2）可以用paddleOCRv3预训练模型直接预测看下效果 3）测试生成数据效果怎么样呢 4）生成的数据和真实数据还是有一些偏差，可以在生成数据集上再改造下

PaddleOCR V3版本的检测和识别模型，断点训练，epoch从1开始

通过Global.checkpoints指定要加载的模型路径进行断点继续训练，参数不需要修改

老师，请问ser+re任务里linking的标注在ppocrlabel的哪里呢？在线等

看看这个[文档](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/PPOCRLabel/README_ch.md#%E8%BF%91%E6%9C%9F%E6%9B%B4%E6%96%B0)介绍哈