Factorized Bilinear Models for Image Recognition
This addresses the problem of limited non-linearity in CNNs for image recognition, offering a novel method with practical efficiency, though it is incremental in building upon existing approaches.
The authors tackled the limitation of linear transformations in CNNs by proposing a Factorized Bilinear layer to model pairwise feature interactions with quadratic terms, achieving superior results on datasets like CIFAR-10, CIFAR-100, and ImageNet compared to state-of-the-art models.
Although Deep Convolutional Neural Networks (CNNs) have liberated their power in various computer vision tasks, the most important components of CNN, convolutional layers and fully connected layers, are still limited to linear transformations. In this paper, we propose a novel Factorized Bilinear (FB) layer to model the pairwise feature interactions by considering the quadratic terms in the transformations. Compared with existing methods that tried to incorporate complex non-linearity structures into CNNs, the factorized parameterization makes our FB layer only require a linear increase of parameters and affordable computational cost. To further reduce the risk of overfitting of the FB layer, a specific remedy called DropFactor is devised during the training process. We also analyze the connection between FB layer and some existing models, and show FB layer is a generalization to them. Finally, we validate the effectiveness of FB layer on several widely adopted datasets including CIFAR-10, CIFAR-100 and ImageNet, and demonstrate superior results compared with various state-of-the-art deep models.