Representation Learning Based on Autoencoder and Deep Adaptive Clustering for Image Clustering

Algorithm 1

	Input: input images ; number of clusters ; the threshold and ; the batch size and the learning rate and ; the balance coefficient
Output: reconstruction images ; clustering labels of input images.
(1)	Pertain a fully convolutional autoencoder;
(2)	for in do
(3)	Extract the images feature in the batch ;
(4)	Compute the similarity and pseudolabel based on (2);
(5)	Calculate the indicator coefficient based on (4);
(6)	Updata , and by minimizing (7);
(7)	end for
(8)	Updata by minimizing (7).
(9)	for in do
(10)	and ;
(11)	end for