Open yyyyzy1 opened 1 year ago
In the first round of training,there was a error''int' object has no attribute' in file 'utils_fit.py',line 87,in fit_one_epoch
Can you describe in more detail the problem you're encountering? Lines 86-89 are mainly used to record log files. You can try to delete it without affecting the training.
In the first round of training,there was a error''int' object has no attribute' in file 'utils_fit.py',line 87,in fit_one_epoch
Can you describe in more detail the problem you're encountering? Lines 86-89 are mainly used to record log files. You can try to delete it without affecting the training.
非常感谢您的回复!我还想请教您一些问题,pre_train模型是放在那个地方使用呢?另外,进行可视化操作的文件是predict_yolo.py文件吗?非常感谢!
In the first round of training,there was a error''int' object has no attribute' in file 'utils_fit.py',line 87,in fit_one_epoch
Can you describe in more detail the problem you're encountering? Lines 86-89 are mainly used to record log files. You can try to delete it without affecting the training.
非常感谢您的回复!我还想请教您一些问题,pre_train模型是放在那个地方使用呢?另外,进行可视化操作的文件是predict_yolo.py文件吗?非常感谢!
预训练模型可以放在任意位置,但是需要修改代码中载入权重时的相对路径。另外predict_yolo.py是单独对目标检测分支的结果进行可视化。 如果还有什么疑问,欢迎提问。
In the first round of training,there was a error''int' object has no attribute' in file 'utils_fit.py',line 87,in fit_one_epoch
Can you describe in more detail the problem you're encountering? Lines 86-89 are mainly used to record log files. You can try to delete it without affecting the training.
非常感谢您的回复!我还想请教您一些问题,pre_train模型是放在那个地方使用呢?另外,进行可视化操作的文件是predict_yolo.py文件吗?非常感谢!
预训练模型可以放在任意位置,但是需要修改代码中载入权重时的相对路径。另外predict_yolo.py是单独对目标检测分支的结果进行可视化。 如果还有什么疑问,欢迎提问。
非常感谢老师您的回复!我现在已经把整体大致跑出来了,但是还不太明白怎么进行论文里gaze estimation(GE)、gaze object prediction(GOD)的可视化,请问老师您的代码中有这部分的可视化操作吗?再次感谢老师!
目前版本中并未包含GE和GOD的可视化代码,如果有需要你可以尝试利用GE部分输出的heatmap作为蒙版进行相关的可视化。
In the first round of training,there was a error''int' object has no attribute' in file 'utils_fit.py',line 87,in fit_one_epoch
Can you describe in more detail the problem you're encountering? Lines 86-89 are mainly used to record log files. You can try to delete it without affecting the training.
非常感谢您的回复!我还想请教您一些问题,pre_train模型是放在那个地方使用呢?另外,进行可视化操作的文件是predict_yolo.py文件吗?非常感谢!
预训练模型可以放在任意位置,但是需要修改代码中载入权重时的相对路径。另外predict_yolo.py是单独对目标检测分支的结果进行可视化。 如果还有什么疑问,欢迎提问。
非常感谢老师您的回复!我现在已经把整体大致跑出来了,但是还不太明白怎么进行论文里gaze estimation(GE)、gaze object prediction(GOD)的可视化,请问老师您的代码中有这部分的可视化操作吗?再次感谢老师!
目前版本中并未包含GE和GOD的可视化代码,你可以尝试利用GE部分输出的heatmap作为蒙版进行相关的可视化操作。我们也会再后续的工作中对开源的代码进行整理完善。
目前版本中并未包含GE和GOD的可视化代码,如果有需要你可以尝试利用GE部分输出的heatmap作为蒙版进行相关的可视化。
In the first round of training,there was a error''int' object has no attribute' in file 'utils_fit.py',line 87,in fit_one_epoch
Can you describe in more detail the problem you're encountering? Lines 86-89 are mainly used to record log files. You can try to delete it without affecting the training.
非常感谢您的回复!我还想请教您一些问题,pre_train模型是放在那个地方使用呢?另外,进行可视化操作的文件是predict_yolo.py文件吗?非常感谢!
预训练模型可以放在任意位置,但是需要修改代码中载入权重时的相对路径。另外predict_yolo.py是单独对目标检测分支的结果进行可视化。 如果还有什么疑问,欢迎提问。
非常感谢老师您的回复!我现在已经把整体大致跑出来了,但是还不太明白怎么进行论文里gaze estimation(GE)、gaze object prediction(GOD)的可视化,请问老师您的代码中有这部分的可视化操作吗?再次感谢老师!
目前版本中并未包含GE和GOD的可视化代码,你可以尝试利用GE部分输出的heatmap作为蒙版进行相关的可视化操作。我们也会再后续的工作中对开源的代码进行整理完善。
老师,您好!非常感谢您的回复,在您写的readme中提到:When you training the model, the index of gaze will be saved in gaze_performence.txt,但是当我train_mode=0的时候并没有这个txt文件生成,只有包含AUC,Dist,Ang的performance.csv文件,没有预测的目标结果txt,请问这是怎么回事呢?
In the first round of training,there was a error''int' object has no attribute' in file 'utils_fit.py',line 87,in fit_one_epoch