Output results of convolution operation with multiple tensors and multiple convolution kernels

import tensorflow as tf


inputValue = tf.constant([
    #1 A tensor 
    [
        #3 That's ok 3 Column 2 depth 
        [[2, 5], [3, 3], [8, 2]],
        [[6, 1], [1, 2], [5, 4]],
        [[7, 9], [2, 3], [-1, 3]]
    ],
    [
        [[1, 3], [2, 1], [3, 2]],
        [[1, 1], [2, 2], [1, 4]],
        [[3, 4], [4, 2], [-1, 1]]
    ]
], tf.float32)

kernels = tf.constant([
    # 2 That's ok 
    [
        # 2 Column 
        [
            # 2 depth 
            [3, 1, -3], [1, -1, 7]
        ],
        [
            [-2, 2, -5], [2, 7, 3]
        ]
    ],
    [
        # 2 Column 
        [
            [-1, 3, 1], [-3, -8, 6]
        ],
        [
            [4, 6, 8], [5, 9, -5]
        ]
    ]
], tf.float32)

validResult = tf.nn.conv2d(inputValue, kernels, [1, 1, 1, 1], "VALID")

print(validResult)

2 individual 3 That's ok 3 Column 2 Depth input and 3 individual 2 That's ok 2 Column 2 Convolution of depth and convolution

The depth of the input quantity must be the same as the depth of the convolution sum

The depth of the output is the number of convolution kernels

The number of output quantity is the number of input quantity

The number of rows and columns of the output quantity is the same as padding and stride of , Convolution operation , In the two dimensions of number and depth stride Take all 1, Row and column stride As the case may be .

Understanding of tensor and convolution , recommend 《 Graphic deep learning and neural network ： From tensor to TensorFlow Realization 》_ Zhang Ping's book