tf.stack()和tf.concat()矩阵拼接

tf.stack(dim,axis,name)

在目标检测代码中用到这个函数，举例说明：

t1 = [[1,2,3],[4,5,6]]
t2 = [[7,8,9],[10,11,12]]
with tf.Session() as sess:
    print(sess.run(tf.stack([t1,t2],axis=0)))
    print('===================================')
    print(sess.run(tf.stack([t1, t2], axis=1)))

输出：

[[[ 1  2  3]
  [ 4  5  6]]

 [[ 7  8  9]
  [10 11 12]]]
===================================
[[[ 1  2  3]
  [ 7  8  9]]

 [[ 4  5  6]
  [10 11 12]]]

tf.concat(concat_dim,axis,name)

t1 = [[1,2,3],[4,5,6]]
t2 = [[7,8,9],[10,11,12]]
with tf.Session() as sess:
    print(sess.run(tf.concat([t1,t2],axis=0)))
    print('===================================')
    print(sess.run(tf.concat([t1, t2], axis=1)))

输出：

[[ 1  2  3]
 [ 4  5  6]
 [ 7  8  9]
 [10 11 12]]
===================================
[[ 1  2  3  7  8  9]
 [ 4  5  6 10 11 12]]