1  tf.cast():    https://blog.csdn.net/dss875914213/article/details/86558407

2 tf.gather :               https://blog.csdn.net/guotong1988/article/details/53172882 

3 tf.boolean_mask:            https://blog.csdn.net/m0_37393514/article/details/81674489

4  tf.where(）:                https://blog.csdn.net/ustbbsy/article/details/79564828

5 tf.gather_nd()       多维上进行索引 https://blog.csdn.net/Cyiano/article/details/76087747

a = tf.Variable([[1,2,3,4,5], [6,7,8,9,10], [11,12,13,14,15]])
index_a = tf.Variable([[0,2], [0,4], [2,2]])

print(sess.run(tf.gather_nd(a, index_a)))

# [ 3 5 13]

6 tf.image.crop_and_resize的使用  ROI Pooling层:https://blog.csdn.net/m0_38024332/article/details/81779544