Research on Chest Disease Recognition Based on Deep Hierarchical Learning Algorithm

Chest X-ray has become one of the most common ways in diagnostic radiology exams, and this technology assists expert radiologists with finding the patients at potential risk of cardiopathy and lung diseases. However, it is still a challenge for expert radiologists to assess thousands of cases in a short period so that deep learning methods are introduced to tackle this problem. Since the diseases have correlations with each other and have hierarchical features, the traditional classification scheme could not achieve a good performance. In order to extract the correlation features among the diseases, some GCN-based models are introduced to combine the features extracted from the images to make prediction. This scheme can work well with the high quality of image features, so backbone with high computation cost plays a vital role in this scheme. However, a fast prediction in diagnostic radiology is also needed especially in case of emergency or region with low computation facilities, so we proposed an efficient convolutional neural network with GCN, which is named SGGCN, to meet the need of efficient computation and considerable accuracy. SGGCN used SGNet-101 as backbone, which is built by ShuffleGhost Block (Huang et al., 2021) to extract features with a low computation cost. In order to make sufficient usage of the information in GCN, a new GCN architecture is designed to combine information from different layers together in GCNM module so that we can utilize various hierarchical features and meanwhile make the GCN scheme faster. The experiment on CheXPert datasets illustrated that SGGCN achieves a considerable performance. Compared with GCN and ResNet-101 (He et al., 2015) backbone (test AUC 0.8080, parameters 4.7M and FLOPs 16.0B), the SGGCN achieves 0.7831 (−3.08%) test AUC with parameters 1.2M (−73.73%) and FLOPs 3.1B (−80.82%), where GCN with MobileNet (Sandler and Howard, 2018) backbone achieves 0.7531 (−6.79%) test AUC with parameters 0.5M (−88.46%) and FLOPs 0.66B (−95.88%).

Centralized Threshold Key Generation Protocol Based on Shamir Secret Sharing and HMAC Authentication

Many group key management protocols have been proposed to manage key generation and distribution of vehicular communication. However, most of them suffer from high communication and computation costs due to the complex elliptic curve and bilinear pairing cryptography. Many shared secret protocols have been proposed using polynomial evaluation and interpolation to solve the previous complexity issues. This paper proposes an efficient centralized threshold shared secret protocol based on the Shamir secret sharing technique and supporting key authentication using Hashed Message Authentication Code Protocol (HMAC). The proposed protocol allows the group manager to generate a master secret key for a group of n vehicles and split this key into secret shares; each share is distributed securely to every group member. t-of-n vehicles must recombine their secret shares and recover the original secret key. The acceptance of the recovered key is based on the correctness of the received HMAC signature to verify the group manager’s identity and ensure the key confidentiality. The proposed protocol is unconditionally secure and unbreakable using infinite computing power as t, or more than t secret shares are required to reconstruct the key. In contrast, attackers with t−1 secret shares cannot leak any information about the original secret key. Moreover, the proposed protocol reduces the computation cost due to using polynomial evaluation to generate the secret key and interpolation to recover the secret key, which is very simple and lightweight compared with the discrete logarithm computation cost in previous protocols. In addition, utilizing a trusted group manager that broadcasts some public information is important for the registered vehicles to reconstruct the key and eliminate secure channels between vehicles. The proposed protocol reduces the communication cost in terms of transmitted messages between vehicles from 2(t−1) messages in previous shared secret protocols to zero messages. Moreover, it reduces the received messages at vehicles from 2t to two messages. At the same time, it allows vehicles to store only a single secret share compared with other shared secret protocols that require storage of t secret shares. The proposed protocol security level outperforms the other shared secret protocols security, as it supports key authentication and confidentiality using HMAC that prevents attackers from compromising or faking the key.

Personalized Route Planning System Based on Driver Preference

At present, most popular route navigation systems only use a few sensed or measured attributes to recommend a route. Yet the optimal route considered by drivers needs be based on multiple objectives and multiple attributes. As a result, these existing systems based on a single or few attributes may fail to meet such drivers’ needs. This work proposes a driver preference-based route planning (DPRP) model. It can recommend an optimal route by considering driver preference. We collect drivers’ preferences, and then provide a set of routes for their choice when they need. Next, we present an integrated algorithm to solve DPRP, which speeds up the search process for recommending the best routes. Its computation cost can be reduced by simplifying a road network and removing invalid sub-routes. Experimental results demonstrate its effectiveness.

Early Prediction of DNN Activation Using Hierarchical Computations

Deep Neural Networks (DNNs) have set state-of-the-art performance numbers in diverse fields of electronics (computer vision, voice recognition), biology, bioinformatics, etc. However, the process of learning (training) from the data and application of the learnt information (inference) process requires huge computational resources. Approximate computing is a common method to reduce computation cost, but it introduces loss in task accuracy, which limits their application. Using an inherent property of Rectified Linear Unit (ReLU), a popular activation function, we propose a mathematical model to perform MAC operation using reduced precision for predicting negative values early. We also propose a method to perform hierarchical computation to achieve the same results as IEEE754 full precision compute. Applying this method on ResNet50 and VGG16 shows that up to 80% of ReLU zeros (which is 50% of all ReLU outputs) can be predicted and detected early by using just 3 out of 23 mantissa bits. This method is equally applicable to other floating-point representations.

Use of the Symmetries in the Study of Vibration Response of a Hollow Cylinder

The paper studies the vibration response of an elastic solid that has geometric symmetries. These determine special properties of the equations of motion of such a system, presented in the case of a cylindrical body (hollow cylinder). The properties of the eigenvalues and eigenmodes of these systems are theoretically established. A validation of these results is made using the finite element method. The use of the obtained results can lead to an easing of the vibration analysis of such a system and, consequently, to the decrease of the cost related to the design and manufacture of such a structure. The properties presented and demonstrated in the paper can simplify the numerical calculation and experimental verifications of such a structure. Serving these symmetries, the computation cost decrease substantially and will depend not in the number of the identical parts.

Anonymous Certificate-Based Inner Product Broadcast Encryption

Broadcast encryption scheme enables a sender distribute the confidential content to a certain set of intended recipients. It has been applied in cloud computing, TV broadcasts, and many other scenarios. Inner product broadcast encryption takes merits of both broadcast encryption and inner product encryption. However, it is crucial to reduce the computation cost and to take the recipient’s privacy into consideration in the inner product broadcast encryption scheme. In order to address these problems, we focus on constructing a secure and practical inner product broadcast encryption scheme in this paper. First, we build an anonymous certificate-based inner product broadcast encryption scheme. Especially, we give the concrete construction and security analysis. Second, compared with the existing inner product broadcast encryption schemes, the proposed scheme has an advantage of anonymity. Security proofs show that the proposed scheme achieves confidentiality and anonymity against adaptive chosen-ciphertext attacks. Finally, we implement the proposed anonymous inner product broadcast encryption scheme and evaluate its performance. Test results show that the proposed scheme supports faster decryption operations and has higher efficiency.

Performance Evaluation of RSA-based Secure Cloud Storage Protocol using OpenStack

In this paper, the implementation of the General Secure Cloud Storage Protocol is carried out and instantiated by a multiplicatively Homomorphic Encryption Scheme (HES). The protocol provides a system for secure storage of data over the cloud, thereby allowing the client to carry out the operational tasks on it efficiently. The work focuses on the execution of five major modules of the protocol. We also evaluate the performance of the protocol with respect to the computation cost of these modules on the basis of different security parameters and datasets by conducting a series of experiments. The cloud was built using OpenStack and the data were outsourced from the client’s system to the cloud to study the security features and performance metrics when adopting the cloud environment.

Modality-aware Style Adaptation for RGB-Infrared Person Re-Identification

RGB-infrared (IR) person re-identification is a challenging task due to the large modality gap between RGB and IR images. Many existing methods bridge the modality gap by style conversion, requiring high-similarity images exchanged by complex CNN structures, like GAN. In this paper, we propose a highly compact modality-aware style adaptation (MSA) framework, which aims to explore more potential relations between RGB and IR modalities by introducing new related modalities. Therefore, the attention is shifted from bridging to filling the modality gap with no requirement on high-quality generated images. To this end, we firstly propose a concise feature-free image generation structure to adapt the original modalities to two new styles that are compatible with both inputs by patch-based pixel redistribution. Secondly, we devise two image style quantification metrics to discriminate styles in image space using luminance and contrast. Thirdly, we design two image-level losses based on the quantified results to guide the style adaptation during an end-to-end four-modality collaborative learning process. Experimental results on two datasets SYSU-MM01 and RegDB show that MSA achieves significant improvements with little extra computation cost and outperforms the state-of-the-art methods.