Search

Abstract Diffusion models are increasingly applied in low‐light image enhancement tasks due to their exceptional capability to model data distributions, but most current methods focus only on the original pixel space and neglect the potential of Fourier frequency information. In this article, SFDiff is proposed, a novel low‐light image enhancement method that integrates Fourier frequency information into the diffusion process. Specifically, Fourier transforms are applied at both the image and feature levels to separately enhance the amplitude and phase components, which restores global illumination degradation and positional information. Then a Spatial‐Frequency Fusion (SFF) block is used to fully integrate and interact with the information across spatial and frequency domains. Since illumination degradation is primarily manifested in the amplitude component, a loss function based on maximum likelihood learning is employed to constrain the amplitude component at each step of the sampling process, ensuring that the reverse process maintains an optimal trajectory. Owing to the streamlined network design and the fact that the Fourier transform requires no extra parameters, SFDiff achieves a reduction in parameters of over 35% compared to several state‐of‐the‐art (SOTA) diffusion models and delivers high‐quality enhancement results on multiple real‐world datasets. The code is available at https://github.com/MrWan001/SFDiff.

Abstract Traffic sign detection is critical for autonomous driving technology. However, accurately detecting traffic signs in complex traffic environments remains challenge despite the widespread use of one‐stage detection algorithms known for their real‐time processing capabilities. In this paper, the authors propose a traffic sign detection method based on YOLO v8. Specifically, this study introduces the Space‐to‐Depth (SPD) module to address missed detections caused by multi‐scale variations of traffic signs in traffic scenes. The SPD module compresses spatial information into depth channels, expanding the receptive field and enhancing the detection capabilities for objects of varying sizes. Furthermore, to address missed detections caused by complex backgrounds such as trees, this paper employs the Select Kernel attention mechanism. This mechanism enables the model to dynamically adjust its focus and more effectively concentrate on key features. Additionally, considering the uneven distribution of training data, the authors adopted the WIoUv3 loss function, which optimizes loss calculation through a weighted approach, thereby improving the model's detection performance across various sizes and frequencies of instances. The proposed methods were validated on the CCTSDB and TT100K datasets. Experimental results demonstrate that the authors’ method achieves substantial improvements of 3.2% and 5.1% on the mAP50 metric compared to YOLOv8s, while maintaining high detection speed, significantly enhancing the overall performance of the detection system. The code for this paper is located at https://github.com/dusongjie/TSD‐YOLO‐Small‐Traffic‐Sign‐Detection‐Based‐on‐Improved‐YOLO‐v8

Abstract The limited of texture details information in low‐resolution facial or eye images presents a challenge for gaze estimation. To address this, FSKT‐GE (feature maps similarity knowledge transfer for low‐resolution gaze estimation) is proposed, a gaze estimation framework consisting of both a high resolution (HR) network and low resolution (LR) network with the identical structure. Rather than mere feature imitation, this issue is addressed by assessing the cosine similarity of feature layers, emphasizing the distribution similarity between the HR and LR networks. This enables the LR network to acquire richer knowledge. This framework utilizes a combination loss function, incorporating cosine similarity measurement, soft loss based on probability distribution difference and gaze direction output, along with a hard loss from the LR network output layer. This approach on low‐resolution datasets derived from Gaze360 and RT‐Gene datasets is validated, demonstrating excellent performance in low‐resolution gaze estimation. Evaluations on low‐resolution images obtained through 2×, 4×, and 8× down‐sampling are conducted on two datasets. On the Gaze360 dataset, the lowest mean angular errors of 10.97°, 11.22°, and 13.61° were achieved, while on the RT‐Gene dataset, the lowest mean angular errors of 6.73°, 6.83°, and 7.75° were obtained.

IntroductionThe surgery of clival chordoma remains one of the most formidable challenges for neurosurgeons because of its location at great depth in the cranium and proximity to critical neurovascular structures. Here, we describe the technique and feasibility of the purely endoscopic far-lateral supracerebellar infratentorial approach (EF-SCITA) for resection of an intradural clival chordoma.Case descriptionA 68-year-old women presented with sudden ptosis on the left side for two weeks. Imaging examinations revealed an upper-middle clival lesion that transgressed dural confines towards the posterior fossa, which was separated from the sphenoid cavity by an intact thin layer of membrane structure in front. For surgery, the EF-SCITA approach via suboccipital craniotomy was attempted for protecting surrounding neurovascular tissue and the membrane barrier under direct vision. The patients were placed in a “head-up” lateral park-bench position. With the endoscopic holder, endoscopic procedures were performed using standard two-hand microsurgical techniques by one surgeon. Tentorium incision allowed a working corridor toward the clival bulge through the crural cistern, without brain traction seen in traditional retrosigmoid approach. Efficient tumor debulking facilitated the exposure of surrounding critical structures, including ipsilateral CN III and superior cerebellar artery above, the brainstem and basilar artery posteriorly, as well as ipsilateral CN VI displaced laterally, and subsequent tumor separation from them. Step-wise tumor resection was performed within dural and bone confines. After significant tumor removal, the pituitary stalk could be visualized anteriorly, together with contralateral internal carotid artery and CN III. Postoperative MRI depicted gross total excision of the lesion. The patient on follow-up at one year had complete recovery of cranial nerve functions, without signs of cerebrospinal fluid rhinorrhea.DiscussionThis technique combines advantages of the posterolateral approach and endoscopy, allowing access to the upper-middle clivus with seemingly low risks of postoperative morbidity. It would be a safe and effective alternative for resection of this rare entity.

The present study addresses a critical issue within the realm of drones: the challenge of distributed constrained optimization. Our research delves into an optimization scenario where the decision variable is confined to a closed convex set. The primary objective is to develop a distributed algorithm capable of tackling this optimization problem. To achieve this, we have crafted distributed algorithms for both balanced graphs and unbalanced graphs, with the method of feasible direction employed to address the considered constraint, and the method of estimating left eigenvector to address the unbalance, incorporating momentum elements. We have demonstrated that the algorithms exhibit linear convergence when the local objective functions are both smooth and strongly convex, and when the step-sizes are appropriately chosen. Additionally, the simulation outcomes validate the efficacy of our distributed algorithms.

With the development of generative model, the cost of facial manipulation and forgery is becoming lower and lower. Fraudulent data has brought numerous hidden threats in politics, privacy, and cybersecurity. Although many methods of face forgery detection focus on the learning of high frequency forgery traces and achieve promising performance, these methods usually learn features in spatial and frequency independently. In order to combine the information of the two domains, a combined spatial and frequency dual stream network is proposed for face forgery detection. Concretely, a cross self-attention (CSA) module is designed to improve frequency feature interaction and fusion at different scales. Moreover, to augment the semantic and contextual information, frequency guided spatial feature extraction module is proposed to extract and reconstruct the spatial information. These two modules deeply mine the forgery traces via a dual-stream collaborative network. Through comprehensive experiments on different datasets, we demonstrate the effectiveness of proposed method for both within and cross datasets.

Abstract Retraction: [Jiazheng Yuan, Zhuang Wang, Cheng Xu, Hongtian Li, Songyin Dai, Hongzhe Liu, Multi‐vehicle group‐aware data protection model based on differential privacy for autonomous sensor networks, IET Circuits, Devices & Systems 2022 (https://doi.org/10.1049/cds2.12140)].The above article from IET Circuits, Devices & Systems, published online on 29 December 2022 in Wiley Online Library (wileyonlinelibrary.com), has been retracted by agreement between the Editor‐in‐Chief, Harry E. Ruda, the Institution of Engineering and Technology (the IET) and John Wiley and Sons Ltd. This article was published as part of a Guest Edited special issue. Following an investigation, the IET and the journal have determined that the article was not reviewed in line with the journal's peer review standards and there is evidence that the peer revie process of the special issue underwent systematic manipulation. Accordingly, we cannot vouch for the integrity or reliability of the content. As such we have taken the decision to retract the article. The authors have been informed of the decision to retract.

Image dehazing is one of the problems that need to be solved urgently in the field of computer vision. In recent years, more and more algorithms have been applied to image dehazing and achieved good results. However, the image after dehazing still has color distortion, contrast and saturation disorder, and other challenges; in order to solve these problems, in this paper, an effective image dehazing method is proposed, which is based on improved color channel transfer and multiexposure image fusion to achieve image dehazing. First, the image is preprocessed using a color channel transfer method based on k-means. Second, gamma correction is introduced on the basis of guided filtering to obtain a series of multiexposure images, and the obtained multiexposure images are fused into a dehazed image through a Laplacian pyramid fusion scheme based on local similarity of adaptive weights. Finally, contrast and saturation corrections are performed on the dehazed image. Experimental verification is carried out on synthetic dehazed images and natural dehazed images, and it is verified that the method proposed is superior to existing dehazed algorithms from both subjective and objective aspects.

The strengthening application of brick concrete recycled coarse aggregate (BCRCA) has become a “troublesome issue.” In this sense, an in-depth study on the strengthening method of BCRCA is needed. In order to improve the strengthening effect of BCRCA and perform the large-scale application of brick concrete waste, the strengthening method of BCRCA was investigated. Briefly, the performance of BCRCA was initially analyzed. Then, the pore characteristics of BCRCA were tested by CT scanning technology. The evaluation index of filling saturation was proposed. Finally, the BCRCA was strengthened by the negative pressure, and the influences of key parameters such as negative pressure, negative pressure time, negative pressure temperature, and stirring times on the strengthening effect were investigated. The test results showed that the best performance can be obtained by using the following parameters: negative pressure from −55 kPa to −80 kPa, the negative pressure time from 6 to 14 min, the negative pressure temperature from 40 to 80°C, and the stirring times from 1 to 4 times. Combined with the influence of negative pressure strengthening process parameters on filling saturation, it can be determined that a negative pressure of −75 kPa, a negative pressure time of 12 min, and a negative pressure temperature of 70°C with two times of stirring were the best negative pressure strengthening parameters. The unconfined compressive strength and indirect tensile strength reached 4.9 and 0.55 MPa, respectively, when the reinforced BCRCA was used for the cement-stabilized macadam. Therefore, the negative pressure strengthening process parameters determined in this study can achieve a good strengthening effect.

Unsupervised domain adaptive person re-identification aims to solve the problem of poor performance caused by transferring an unlabeled target domain from the labeled source domain in the re-identification task. The clustering pseudo-labels method in unsupervised learning is widely used in unsupervised adaptive person re-identification tasks, and it maintains state-of-the-art performance. However, pseudo-labels obtained through clustering often have much noise, and the use of a single network model structure and a single clustering algorithm can easily cause model learning to stagnate, making the model not generalizable. To solve this problem, this paper proposes an asymmetric mutual mean-teaching method for unsupervised adaptive person re-identification. In terms of feature extraction, two asymmetric network models with different structures are used for mutual mean-teaching on the target domain, making the features extracted by the network more robust. In terms of feature clustering, two clustering methods are used for mutual teaching to dynamically update the centroid of clusters to improve the confidence of clustering pseudo-labels. Finally, the triplet loss is improved based on the updated cluster centroid to improve the clustering effect. The proposed method is used to perform a large number of verification experiments on three public datasets. The experimental results show that the proposed method has better accuracy than other unsupervised person re-identification based on clustering pseudo-labels.

Heatmap-based traditional approaches for estimating human pose usually suffer from drawbacks such as high network complexity or suboptimal accuracy. Focusing on the issue of multi-person pose estimation without heatmaps, this paper proposes an end-to-end, lightweight human pose estimation network using a multi-scale coordinate attention mechanism based on the Yolo-Pose network to improve the overall network performance while ensuring the network is lightweight. Specifically, the lightweight network GhostNet was first integrated into the backbone to alleviate the problem of model redundancy and produce a significant number of effective feature maps. Then, by combining the coordinate attention mechanism, the sensitivity of our proposed network to direction and location perception was enhanced. Finally, the BiFPN module was fused to balance the feature information of different scales and further improve the expression ability of convolutional features. Experiments on the COCO 2017 dataset showed that, compared with the baseline method YOLO-Pose, the average accuracy of the proposed network on the COCO 2017 validation dataset was improved by 4.8% while minimizing the amount of network parameters and calculations. The experimental results demonstrated that our proposed method can improve the detection accuracy of human pose estimation while ensuring that the model is lightweight.

Spark-ignited natural gas engines have received increasing attention in the heavy-duty market due to their low cost and reliability advantages. However, there are still some issues with natural gas engines retrofitted from 10 to 15 L diesel engines, which is a valuable medium-term goal for the automotive industry. In this work, the effect of intake port structure on the performance of a spark-ignited heavy-duty natural gas engine was investigated by multidimensional numerical simulations. A newly designed intake port was proposed, with strengthened in-cylinder turbulent kinetic energy and homogeneous air-fuel mixtures. Bench tests show that the proposed intake port has impressive thermal efficiency, cycle variation, and acceptable emissions performance. The effective thermal efficiency improves from 41.0% to 41.4%, and the cycle variation is 36% lower than traditional schemes. However, with the accelerated flame propagation, the in-cylinder temperature and NOx emission of the mixed-flow port increase while the CO emission decreases. In summary, a proper balance of in-cylinder swirl and tumble flow can significantly affect the economy and stability of natural gas engines. The proposed structure solves the inherent problems of slow natural gas flame propagation and harmful cyclic variations.

With the development of science and technology, artificial intelligence has been widely used in the transportation field, and research on the symmetry of artificial intelligence has become increasingly more in-depth. Traffic sign detection based on deep learning has the problems of different target shapes and high variability in the number of targets between different labels. To solve these problems from a lack of symmetry, the idea of applying the concept of balanced data and the deformable positioning region to a target recognition network is proposed. The research is based on the improvement of the Libra R-CNN. Aiming at the problem that the difficult-to-distinguish target in target detection has a high impact on detection, the idea of generating increasingly more diverse indistinguishable samples during training is proposed to improve the detection accuracy, which is verified by experiments. The experiment is carried out on the MS COCO 2017 and traffic sign datasets. The improved Libra R-CNN is 3 percentage points better than the unimproved Libra R-CNN's mean Average Precision (mAP). A large number of comparative experimental results show that the improved network is effective.

Multi-label aerial scene image classification is a long-standing and challenging research problem in the remote sensing field. As land cover objects usually co-exist in an aerial scene image, modeling label dependencies is a compelling approach to improve the performance. Previous methods generally directly model the label dependencies among all the categories in the target dataset. However, most of the semantic features extracted from an image are relevant to the existing objects, making the dependencies among the nonexistant categories unable to be effectively evaluated. These redundant label dependencies may bring noise and further decrease the performance of classification. To solve this problem, we propose S-MAT, a Semantic-driven Masked Attention Transformer for multi-label aerial scene image classification. S-MAT adopts a Masked Attention Transformer (MAT) to capture the correlations among the label embeddings constructed by a Semantic Disentanglement Module (SDM). Moreover, the proposed masked attention in MAT can filter out the redundant dependencies and enhance the robustness of the model. As a result, the proposed method can explicitly and accurately capture the label dependencies. Therefore, our method achieves CF1s of 89.21%, 90.90%, and 88.31% on three multi-label aerial scene image classification benchmark datasets: UC-Merced Multi-label, AID Multi-label, and MLRSNet, respectively. In addition, extensive ablation studies and empirical analysis are provided to demonstrate the effectiveness of the essential components of our method under different factors.

To study the mechanical behavior and influence factors of the reinforced retaining wall under the static load, numerical simulation of the reinforced retaining wall is conducted by finite element analysis, and its mechanical behavior and influencing methods are studied in accordance with relevant theories. The results showed that the properties of back fill, reinforced spacing, reinforced stiffness, reinforced length, and panel stiffness all affect the mechanical behavior of retaining walls. According to the example calculations of different wall heights, the distribution of panel horizontal displacement and maximum tensile stress are analyzed. The gravel with good gradation has better durability and can reduce the amount of reinforcing steel; with the decrease of the reinforcement spacing, the deformation of the wall panel will become smaller, and the reinforcement effect will be improved; the length of reinforcement is not the longer the better, and the deformation of wall panel can be minimized at the suitable length; the larger the elastic modulus of the wall panel, the smaller the deformation of the wall panel will be.

Abstract Semantic segmentation of an image scene provides semantic information of image regions while less information of objects. In this paper, we propose a method of hierarchical semantic segmentation, including scene level and object level, which aims at labeling both scene regions and objects in an image. In the scene level, we use a feature-based MRF model to recognize the scene categories. The raw probability for each category is predicted via a one-vs-all classification mode. The features and raw probability of superpixels are embedded into the MRF model. With the graph-cut inference, we get the raw scene-level labeling result. In the object level, we use a constraint-based geodesic propagation to get object segmentation. The category and appearance features are utilized as the prior constraints to guide the direction of object label propagation. In this hierarchical model, the scene-level labeling and the object-level labeling have a mutual relationship, which regions and objects are optimized interactively. The experimental results on two datasets show the well performance of our method.

Lane marks on roads are among the most important items of road scene information in the process of autonomous driving, and lane-mark extraction based on visual cognitive computing is one of the most important components of advanced driving assistance systems in intelligent transportation system. Onboard cameras mounted on the front of autonomous vehicles capture road scene images from which lane marks are extracted. This paper proposes a new lane-mark extraction algorithm with four major parts. First, this paper handles the road images captured from onboard cameras by grayscale and fast median filter. Then, we exploit the characteristics of lane marks in road images as constraints to propose lane-features filter based on multi-constraints used to extract lane marks. Then, a clustering algorithm based on the double point removal of a p-least squares algorithm is proposed to cluster features, and recursive dichotomy algorithm is used to fit the candidate lane marks. Finally, we carry out verification and optimization on candidate lane marks to obtain more accurate and stable extraction results. In our experiment, we divide the common complex road scenes into four categories. The results show that the proposed method can robustly extract lane marks under various complex real conditions. This paper also proposes forward a method to evaluate the results of lane-mark extraction, and partial test results are evaluated.

The vehicle to everything (V2X) requires the real-time integration of all kinds of information on roads, pedestrians, the environment, and vehicles themselves. This information also needs to be shared with other vehicles. The effective integration of information and the strong privacy protection are the key restrictions on the development of the V2X. The previous privacy protection model has mainly focused on the centralized network, and there were problems with the centralized gateway and single-point decision, which were not suitable for the decentralized scenario. Therefore, this paper proposes a remote attestation security model based on a privacy-preserving blockchain. The overall model involves two core steps. First, the vehicle provides the network with an evidence of a credible identity and integrity. Secured, the vehicles in the network calculate the nodes to make their respective decisions, and the accounting nodes summarize the sub-conclusions, form the final results, and write them into data blocks. The analysis shows that it possesses the security features of decentralization, traceability, anonymity, irreplaceability, and high efficiency. The model framework, core block chain structure, and protocol process are described in detail. The experimental results based on a realistic infrastructure are presented. These experimental results demonstrate that our scheme can effectively enhance the security of the communications of intelligent vehicles in the V2X.

The Internet of Energy (IoE) provides an effective networking technology for distributed green energy, which allows the connection of energy anywhere at any time. As an important part of the IoE, electric vehicles (EVs), and charging pile management are of great significance to the development of the IoE industry. Previous work has mainly focused on network performance optimization for its management, and few studies have considered the security of the management between EVs and charging piles. Therefore, this paper proposes a decentralized security model based on the lightning network and smart contract in the blockchain ecosystem; this proposed model is called the lightning network and smart contract (LNSC). The overall model involves registration, scheduling, authentication, and charging phases. The new proposed security model can be easily integrated with current scheduling mechanisms to enhance the security of trading between EVs and charging piles. Experimental results according to a realistic infrastructure are presented in this paper. These experimental results demonstrate that our scheme can effectively enhance vehicle security. Different performances of LNSC-based scheduling strategies are also presented.

Background: The geographic distribution of osteochondroma (OC) varies greatly around the world. There has been no recent report on OC in a large Chinese population. The aim of this study was to characterize OC by an epidemiological analysis of the clinical data from one medical institution in South China. Methods: We searched medical electronic records from January 2001 to January 2016 in one large hospital in South China to identify patients with a definite diagnosis of OC. Their epidemiological data were collected and analyzed statistically, including gender, tumor site, age at first diagnosis and symptoms, local recurrence and malignant transformation. Differences between genders and between solitary osteochondroma (SO) and multiple osteochondroma (MO) were particularly analyzed. Results: A total of 431 OC patients (291 males and 140 females; 329 SOs and 102 MOs) were identified. The gender ratio was 2.08 with a male predominance. OCs were mostly located around the knee (250 cases). 280 patients were in their 0s to 20s upon first diagnosis. The average age at the time of first diagnosis was 20.63 years for all, 18.47 years for males and 25.11 years for females (P=0.000). OC recurred locally in 35 patients (15 SOs and 20 MOs), with a significant difference between SO/MO (P=0.000) but not between genders (P=0.100). The average interval from the primary surgery to local or malignant recurrence was 37.41 months. Malignant transformation was found in 5 patients (4 males and 1 female), showing no gender difference (P=0.549). Conclusions: OC may have a male predominance in Chinese population. It mostly occurred at 0–20 years of age and around the knee. Upon the first diagnosis of OC, the males tended to be younger than the females, and so did the MO patients than the SO ones. In addition, MO had a higher incidence of local recurrence. Intervals from primary surgery to local recurrence or malignant transformation in MO patients were longer than in SO patients.

Objective: To observe the clinical effect of zoledronic acid (ZA) in patients with cervical spondylosis and osteoporosis after anterior cervical discectomy and fusion (ACDF) surgery. Methods: All selected patients were divided into the study group and the control group according to the sequence of surgery time. In the study group, 5 mg (100 ml) of ZA was applied intravenously as intervention on the 5 day after ACDF surgery. Patients were followed up regularly after surgery. Results: Forty-three cases completed the follow-ups (21/22), the neck disability index (NDI) score significantly decreased at the 3rd month after surgery in both groups, it came to a plateau at the 6th month after surgery, and it had some rise at the 12th month after surgery, but the NDI score was lower in the study group at the 12th month after surgery ( p < 0.05). C-telopeptide of type I collagen (CTX) and amino terminal propeptide of type I procollagen of bone turnover markers in the study group showed a downward trend after surgery, among which CTX decreased significantly and no significant changes in the control group. At the 12th month after surgery, the bone mineral density of lumbar spine area in the study group was significantly improved ( p < 0.05). During the 3rd and the 6th month follow-up after surgery, 1/2 osteogenesis rate of the study group was significantly higher ( p < 0.05). However, all patients in the two groups obtained stable fusion at the 12th month follow-up after surgery. Conclusions: For patients with cervical spondylosis and osteoporosis, intravenous application of ZA after ACDF surgery can shorten the time of cervical osteogenesis, promote local bone graft fusion, and improve the postoperative clinical effect to some extent.

Instance segmentation in aerial images is of great significance for remote sensing applications, and it is inherently more challenging because of cluttered background, extremely dense and small objects, and objects with arbitrary orientations. Besides, current mainstream CNN-based methods often suffer from the trade-off between labeling cost and performance. To address these problems, we present a pipeline of hybrid supervision. In the pipeline, we design an ancillary segmentation model with the bounding box attention module and bounding box filter module. It is able to generate accurate pseudo pixel-wise labels from real-world aerial images for training any instance segmentation models. Specifically, bounding box attention module can effectively suppress the noise in cluttered background and improve the capability of segmenting small objects. Bounding box filter module works as a filter which removes the false positives caused by cluttered background and densely distributed objects. Our ancillary segmentation model can locate object pixel-wisely instead of relying on horizontal bounding box prediction, which has better adaptability to arbitrary oriented objects. Furthermore, oriented bounding box labels are utilized for handling arbitrary oriented objects. Experiments on iSAID dataset show that the proposed method can achieve comparable performance (32.1 AP) to fully supervised methods (33.9 AP), which is obviously higher than weakly supervised setting (26.5 AP), when using only 10% pixel-wise labels.

The popularity of social networks has brought the rapid growth of social images which have become an increasingly important image type. One of the most obvious attributes of social images is the tag. However, the sate-of-the-art methods fail to fully exploit the tag information for saliency detection. Thus this paper focuses on salient region detection of social images using both image appearance features and image tag cues. First, a deep convolution neural network is built, which considers both appearance features and tag features. Second, tag neighbor and appearance neighbor based saliency aggregation terms are added to the saliency model to enhance salient regions. The aggregation method is dependent on individual images and considers the performance gaps appropriately. Finally, we also have constructed a new large dataset of challenging social images and pixel-wise saliency annotations to promote further researches and evaluations of visual saliency models. Extensive experiments show that the proposed method performs well on not only the new dataset but also several state-of-the-art saliency datasets.

As one of the most popular statistical and machine learning models, logistic regression with regularization has found wide adoption in biomedicine, social sciences, information technology, and so on. These domains often involve data of human subjects that are contingent upon strict privacy regulations. Concerns over data privacy make it increasingly difficult to coordinate and conduct large-scale collaborative studies, which typically rely on cross-institution data sharing and joint analysis. Our work here focuses on safeguarding regularized logistic regression, a widely-used statistical model while at the same time has not been investigated from a data security and privacy perspective. We consider a common use scenario of multi-institution collaborative studies, such as in the form of research consortia or networks as widely seen in genetics, epidemiology, social sciences, etc. To make our privacy-enhancing solution practical, we demonstrate a non-conventional and computationally efficient method leveraging distributing computing and strong cryptography to provide comprehensive protection over individual-level and summary data. Extensive empirical evaluations on several studies validate the privacy guarantee, efficiency and scalability of our proposal. We also discuss the practical implications of our solution for large-scale studies and applications from various disciplines, including genetic and biomedical studies, smart grid, network analysis, etc.