In short, 3D convolution operations apply a 3D filter to the input data along the three directions, which are x, y, and z. This operation creates a stacked list of 3D feature maps. The shape of the output is similar to the shape of a cube or a cuboid. The following image illustrates a 3D convolution operation. The highlighted part of the left cube is the input data. The kernel is in the middle, with a shape of (3, 3, 3). The block on the right-hand is the output of the convolution operation:

Now that we have a basic understanding of 3D convolutions, let's continue looking at the architecture of a 3D-GAN.