crnn.pytorch.tensorrt.chinese icon indicating copy to clipboard operation
crnn.pytorch.tensorrt.chinese copied to clipboard

A Chinese characters recognition repository with tensorrt format supported based on CRNN_Chinese_Characters_Rec and TensorRTx.

Results 6 crnn.pytorch.tensorrt.chinese issues
Sort by recently updated
recently updated
newest added
trafficstars

使用了360CC的300万样本和它自带的5990的字符, w_pad方式, 开启了image_aug, 训练了5个epoch, 约12h, 显示正确率已经到97%, loss 0.02左右, 测试360cc中的样本检测效果比较好, 但是自己的30x160的文本图, 检测结果几乎不正确 使用您提供的权重, 结果也比较好, 我推测是我使用的360cc样本的训练方式问题, 请问您是基于什么样本训练的, 或者有什么建议, 谢谢!

CRNN的输入是把图片的高度放缩到32,如果按照相同的比例将宽放缩,此时不同的图片的宽度是不一样的,这种情况应该如何训练模型?

Hello, wheni am trying to start training sequence it results with that (cannot import name '_CF' from 'lib.dataset._cf') error. When i checked lib/dataset folder there are other .py documents but...

如果想转换成trt支持的fp16或者Int8类型,应该如何修改。

Building engine, please wait for a while... [04/16/2021-10:54:24] [E] [TRT] ../builder/cudnnBuilderWeightConverters.cpp (555) - Misc Error in operator(): 1 (Weights are outside of fp16 range. A possible fix is to retrain...

我试着重新make的时候报了这样的错,大佬知道是哪里的错误吗? wzp@i-5DB1BBD8:~/project/trt/crnn.pytorch.tensorrt.chinese/crnn_trt/build$ make Scanning dependencies of target crnn [ 50%] Building CXX object CMakeFiles/crnn.dir/crnn.cpp.o /home/wzp/project/trt/crnn.pytorch.tensorrt.chinese/crnn_trt/crnn.cpp:5:21: fatal error: NvInfer.h: No such file or directory #include "NvInfer.h" ^ compilation terminated. CMakeFiles/crnn.dir/build.make:62: recipe...