Robust Classification with Convolutional Prototype Learning

Abstract

Convolutional neural networks (CNNs) have been widely used for image classification. Despite its high accuracies, CNN has been shown to be easily fooled by some adversarial examples, indicating that CNN is not robust enough for pattern classification. In this paper, we argue that the lack of robustness for CNN is caused by the softmax layer, which is a totally discriminative model and based on the assumption of closed world (i.e., with a fixed number of categories). To improve the robustness, we propose a novel learning framework called convolutional prototype learning (CPL). The advantage of using prototypes is that it can well handle the open world recognition problem and therefore improve the robustness. Under the framework of CPL, we design multiple classification criteria to train the network. Moreover, a prototype loss (PL) is proposed as a regularization to improve the intra-class compactness of the feature representation, which can be viewed as a generative model based on the Gaussian assumption of different classes. Experiments on several datasets demonstrate that CPL can achieve comparable or even better results than traditional CNN, and from the robustness perspective, CPL shows great advantages for both the rejection and incremental category learning tasks.

GO Loss: A Gaussian Distribution‐Based Orthogonal Decomposition Loss for Classification

“…Remark Acc. (%) Soft-max -99.28 Center loss [27] λ � 0.1 99.62 Ring loss [23] λ � 0.1 99.58 LGM loss [28] α � 1 99.36 GCPL loss [37] λ � 0.1 99.41 SL [41] η � 0.0 99.32 GO loss λ � 0.1 99.66 ± 0.03 6 Complexity e dataset contains 10 categories of fashion products and is divided into 60, 000 training samples and 10,000 testing samples. We adopt the same network and training parameters with MNIST.…”

“…Regularizing the extracted features or adding regularization terms makes the features of the same class compact and the features of different classes separated. Based on this, several loss functions for classification have been studied from the perspective of redesigning clusters [35,36], such as GCPL loss [37] and Structure-aware loss [38].…”

“…We train the network with different loss functions, namely, soft-max loss, center loss [27], ring loss with softmax [23], LGM loss [28], GCPL loss [37], and SL [41]. In the aforementioned methods, center loss, ring loss, LGM loss, and GCPL loss optimize the loss function from the perspective of intraclass compactness and interclass separation of features.…”

P-ODN: Prototype-based Open Deep Network for Open Set Recognition

“…where S is the batch size, f i is the feature of the ith sample in the batch, and label i is the ground truth. Second, the prototype loss, which is firstly proposed by 16 , is modified to apply in our framework to train the prototypes of knowns. Third, we propose the prototype radius loss, which guides the model to learn the radius scope of each known category.…”

“…Most recently, the prototype learning was introduced to improve the robustness of CNNs. Yang et al proposed the CPL to improve the robustness by using prototypes and proposed the PL (prototype loss) to improve the intra-class compactness and inter-class distance of the feature representation in the work 16 . Yang et al also introduced a method to handle open set recognition problem by using prototypes in their paper.…”