关于keras-yolov3-deepsort

因科研需要，本人最近在跑keras-yolov3代码，本以为很顺利的git一段gihub上的代码简单调试即可。还是我太年轻了，万事开头难啊。下面是我遇到的一些问题，希望可以给你一点帮助，一点信心。科研道路坎坷，贵在坚持啊~

问题一：AttributeError: module 'keras.backend' has no attribute 'get_session'

解决办法：

`from keras import backend as K`

改成：

`from tensorflow.compat.v1.keras import backend as K

问题二：ValueError: Subshape must have computed start >= end since stride is negative, but is 0 and 2 (computed from start 0 and end 9223372036854775807 over shape with rank 2 and stride-1)

解决办法：找到yolo3/model.py

box_xy = (K.sigmoid(feats[..., :2]) + grid) / K.cast(grid_shape[::-1], K.dtype(feats))

box_wh = K.exp(feats[..., 2:4]) * anchors_tensor / K.cast(input_shape[::-1], K.dtype(feats))

改成：

box_xy = (K.sigmoid(feats[..., :2]) + grid) / K.cast(grid_shape[...,::-1], K.dtype(feats))

box_wh = K.exp(feats[..., 2:4]) * anchors_tensor / K.cast(input_shape[...,::-1], K.dtype(feats))

炸眼一看可能看不出，就是把k.cast()里的改成k.cast(×××[…,::-1])

参考网址

[https://stackoverflow.com/questions/57558476/training-a-keras-model-yields-multiple-optimizer-errors]
https://www.cnblogs.com/sherrydatascience/p/13584237.html