With increasing markets for fingerprint authentication, there are also increasing concerns about spoofs or synthetically produced fingerprint identifications that can bypass the authentication process. In this Letter, the authors introduce a new convolutional neural networks (CNNs) architecture for fingerprint liveness detection problem that can provide a more robust framework for network training and detection than previous methods. The proposed method employs squared regression error for each receptive field without the usage of the fully connected layer. Such structure provides following advantages from the previous liveness fingerprint CNN. First, unlike the previous techniques which rely on the pre‐trained features, the proposed CNN can be trained directly from fingerprints as the loss is minimised for each receptive field. Second, in contrast to the cross‐entropy layer, the squared error layer allows them to set up a threshold value that can control the acceptable level of false positives or false negatives. Third, the absence of a fully connected layer allows them to crop the input fingerprints such that a trade‐off between accuracy and computation time can be made without the negative effects of re‐scaling. The proposed CNN is shown to provide higher accuracy for three out of four datasets when evaluated against the state‐of‐the‐art method.

In this paper, we propose an efficient knowledge distillation method to train light networks using heavy networks for semantic segmentation. Most semantic segmentation networks that exhibit good accuracy are based on computationally expensive networks. These networks are not suitable for mobile applications using vision sensors, because computational resources are limited in these environments. In this view, knowledge distillation, which transfers knowledge from heavy networks acting as teachers to light networks as students, is suitable methodology. Although previous knowledge distillation approaches have been proven to improve the performance of student networks, most methods have some limitations. First, they tend to use only the spatial correlation of feature maps and ignore the relational information of their channels. Second, they can transfer false knowledge when the results of the teacher networks are not perfect. To address these two problems, we propose two loss functions: a channel and spatial correlation (CSC) loss function and an adaptive cross entropy (ACE) loss function. The former computes the full relationship of both the channel and spatial information in the feature map, and the latter adaptively exploits one-hot encodings using the ground truth labels and the probability maps predicted by the teacher network. To evaluate our method, we conduct experiments on scene parsing datasets: Cityscapes and Camvid. Our method presents significantly better performance than previous methods.

Fingerprints are known to be easily synthesized to trick identification systems. In this paper, we propose a new method that incorporates template fingerprints stored for identification in the liveness detection system. The fingerprint identification platform must have a list of template fingerprints stored for matching with new probe fingerprints trying to access the system. Thus, instead of simply detecting the liveness of the probe fingerprints, the proposed approach uses the matching template fingerprints along with probe fingerprints through convolutional neural networks to make the liveness decision, which comprises two sequential convolutional neural networks for classification. The proposed method can be built on the top of existing liveness detection methods to increase accuracy without a significant increase in computation time. The evaluation over the LivDet dataset shows that the proposed fingerprint liveness detection method is able to obtain state-of-the-art accuracy.

To achieve high performance, most deep convolutional neural networks (DCNNs) require a significant amount of training data with ground truth labels. However, creating ground-truth labels for semantic segmentation requires more time, human effort, and cost compared with other tasks such as classification and object detection, because the ground-truth label of every pixel in an image is required. Hence, it is practically demanding to train DCNNs using a limited amount of training data for semantic segmentation. Generally, training DCNNs using a limited amount of data is problematic as it easily results in a decrease in the accuracy of the networks because of overfitting to the training data. Here, we propose a new regularization method called pixel-wise adaptive label smoothing (PALS) via self-knowledge distillation to stably train semantic segmentation networks in a practical situation, in which only a limited amount of training data is available. To mitigate the problem caused by limited training data, our method fully utilizes the internal statistics of pixels within an input image. Consequently, the proposed method generates a pixel-wise aggregated probability distribution using a similarity matrix that encodes the affinities between all pairs of pixels. To further increase the accuracy, we add one-hot encoded distributions with ground-truth labels to these aggregated distributions, and obtain our final soft labels. We demonstrate the effectiveness of our method for the Cityscapes dataset and the Pascal VOC2012 dataset using limited amounts of training data, such as 10%, 30%, 50%, and 100%. Based on various quantitative and qualitative comparisons, our method demonstrates more accurate results compared with previous methods. Specifically, for the Cityscapes test set, our method achieved mIoU improvements of 0.076%, 1.848%, 1.137%, and 1.063% for 10%, 30%, 50%, and 100% training data, respectively, compared with the method of the cross-entropy loss using one-hot encoding with ground truth labels.

Previous face deblurring methods have utilized semantic segmentation maps as prior knowledge. Most of these methods generated the segmentation map from a blurred facial image, and restore it using the map in a sequential manner. However, the accuracy of the segmentation affects the restoration performance. Generally, it is difficult to obtain an accurate segmentation map from a blurred image. Instead of sequential methods, we propose an efficient method that learns the flows of facial component restoration without performing segmentation. To this end, we propose a multi-semantic progressive learning (MSPL) framework that progressively restores the entire face image starting from the facial components such as the skin, followed by the hair, and the inner parts (eyes, nose, and mouth). Furthermore, we propose a discriminator that observes the reconstruction-flow of the generator. In addition, we present new test datasets to facilitate the comparison of face deblurring methods. Various experiments demonstrate that the proposed MSPL framework achieves higher performance in facial image deblurring compared to the existing methods, both qualitatively and quantitatively. Our code, trained model and data are available at https://github.com/dolphin0104/MSPL-GAN.