DeepLearningSlideCaptcha icon indicating copy to clipboard operation
DeepLearningSlideCaptcha copied to clipboard

Deep Learning Slide Captcha

Results 3 DeepLearningSlideCaptcha issues
Sort by recently updated
recently updated
newest added

自己训练的模型, 总共用了200张图片训练的 1. 训练完后,看着有一个固定偏差,貌似从中心位置开始画的框 2. 使用模型后的图片变小了,原来 536 300, 变成496 277 3. 横坐标基本都偏离40(正好是原图片横坐标变小的尺寸),纵坐标变小30-35(原图变小23) 4. 打印出来的结果,放在原图上是正确的。 ![enter description here](https://gitee.com/margan/pictures/raw/master/小书匠/1636628731058.png) ![enter description here](https://gitee.com/margan/pictures/raw/master/小书匠/1636628760236.png)

$ sh train.sh 2022-09-02 10:18:02.992340: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cudart64_110.dll 2022-09-02 10:18:04.602552: I tensorflow/compiler/jit/xla_cpu_device.cc:41] Not creating XLA devices, tf_xla_enable_xla_devices not set 2022-09-02 10:18:04.602591: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic...

训练出来的模型有很多个,该选用哪一个呀?也没有生成最终的模型