Stochastic Channel Decorrelation Network and Its Application to Visual Tracking
This addresses the problem of overfitting and inefficiency in CNNs for computer vision practitioners, but it is incremental as it builds on existing methods like Siamese networks.
The paper tackles parameter redundancy in deep CNNs caused by feature correlations between channels by proposing a stochastic channel decorrelation (SCD) block, and demonstrates its effectiveness through a visual tracking algorithm based on Fully-Convolutional Siamese Networks.
Deep convolutional neural networks (CNNs) have dominated many computer vision domains because of their great power to extract good features automatically. However, many deep CNNs-based computer vison tasks suffer from lack of training data while there are millions of parameters in the deep models. Obviously, these two biphase violation facts will result in parameter redundancy of many poorly designed deep CNNs. Therefore, we look deep into the existing CNNs and find that the redundancy of network parameters comes from the correlation between features in different channels within a convolutional layer. To solve this problem, we propose the stochastic channel decorrelation (SCD) block which, in every iteration, randomly selects multiple pairs of channels within a convolutional layer and calculates their normalized cross correlation (NCC). Then a squared max-margin loss is proposed as the objective of SCD to suppress correlation and keep diversity between channels explicitly. The proposed SCD is very flexible and can be applied to any current existing CNN models simply. Based on the SCD and the Fully-Convolutional Siamese Networks, we proposed a visual tracking algorithm to verify the effectiveness of SCD.