Scribble supervision is increasingly vital in cardiac MRI segmentation due to its low annotation cost. However, challenges still exist due to limited supervision signals and the complexity of cardiac MRI. We propose SAM-GA, a SAM-guided segmentation training framework, to address these challenges. SAM generates pseudo-labels to supervise segmentation networks, addressing limited supervision signals challenge. Meanwhile, we propose three strategies, entropy-based point prompting, multi-class fusion, and cross-pseudo-label fine-tune, to optimize these pseudo-labels and adapt SAM for cardiac MRI. Additionally, we design a dual-branch network, Grouped Aggregation Network(GANet), to handle the complexity of cardiac MRI challenge. GANet improves segmentation consistency and accuracy by guiding each branch to focus on different granularity information and combining their outputs. Our comparative experiments on two cardiac benchmark datasets demonstrate that our method achieves state-of-the-art performance, validating its effectiveness in addressing the challenges of weakly supervised cardiac MRI segmentation. The code is available on GitHub.

Retinal OCT anomaly detection is an important research field of ocular disease diagnostics. Reconstruction-based methods utilizing only normal samples are the dominant approach in the field. These studies have yielded significant results. However, two issues remain: (1) Due to the strong generalization of the model, abnormal regions can also be reconstructed correctly, leading to small reconstruction error that affect classification accuracy. (2) The reconstruction-based method only focuses on the internal features of the individual samples, unable to model the large differences among normal samples. In this paper, we propose a model named SR-AD, including a Difference-Awareness Module (DAM) and a Relational Consistency Learning module (RCL). DAM designs a masking strategy that reduces the exposure of abnormal information to avoid good reconstruction of abnormal regions, thus enlarging the abnormal image reconstruction error. RCL reconstructs the relationships between normal samples to model the characteristics of normal samples comprehensively from a global perspective. Experiments on the SpectralisOCT dataset validate that our model achieves advanced performance, reaching 99.07% in AUC.

In autonomous driving, lane changing plays a crucial role in data augmentation. By synthesizing new viewpoints during lane changing, we can effectively increase the diversity of training data, thereby improving the model’s generalization capability in different driving scenarios. However, autonomous driving data is often collected from a single lane, leading to sparse viewpoints, which makes it challenging to synthesize lane changes using existing methods like NeRF and 3D Gaussian Splatting, as they suffer from significant rendering quality degradation, including road texture blurring and distortion, especially with large viewpoint shifts. To address this issue, the LaneGS method is proposed, aiming to optimize novel view synthesis for road regions. Specifically, two key modules are designed: the point cloud normal constraint module and the virtual view regularization module. The point cloud normal constraint module leverages prior knowledge of point cloud normals to guide the orientation and shape of Gaussian ellipsoids, ensuring alignment with the road’s geometry, reducing overfitting, and enhancing multi-view consistency. Furthermore, to further improve the rendering quality of new views after lane changing, the virtual view regularization module incorporates information from the new viewpoint, effectively guiding the optimization process. Qualitative and quantitative comparisons demonstrate the effectiveness of our method in synthesizing new views after lane changing. Our method achieves high-quality lane change view synthesis with a 3-meter horizontal displacement on the Waymo, KITTI, and Virtual KITTI 2 datasets. Additionally, our optimization improves the FID score by 13.40% while maintaining the reconstruction quality of the original views.

Removing dynamic objects and reconstructing high-quality background are crucial for enhancing model editability and scene diversity in autonomous driving. Existing methods face challenges with the holes due to the persistent occlusion of the street background by dynamic objects, limiting flexible scene editing. To address this, we propose a framework, SBR-GS, focusing on removing dynamic objects and improving the reconstruction performance of the static background. Specifically, Inpaint Anything (IA) is integrated to resolve hole issues from continuous occlusions. A texture consistency module is designed to optimize the texture continuity between occluded and visible regions. To mitigate artifacts from the newly generated background lacking depth supervision, a depth estimation network is incorporated to enhance background depth in occluded areas. Additionally, a pixel-aware gradient density control method is introduced to improve overall reconstruction quality, addressing clarity issues due to insufficient initial point clouds. Experimental results indicate that this method achieves a 22.05% improvement in the FID metric on the Waymo dataset and a 26.67% improvement on the KITTI dataset compared to state-of-the-art methods. This approach not only enhances static background reconstruction quality but also supports the development of highly editable models in autonomous driving contexts.

Since autonomous driving requires high-precision object detection in real-time, and multi-line LiDAR generates huge point clouds, developing a lightweight 3D detector is crucial. The high sparsity and unstructured nature of point clouds require transforming raw data into a structured format for effective feature extraction. Nevertheless, despite the decrease in computational complexity achieved through the transformation, the resulting structure exhibits high sparsity. Consequently, using conventional neural networks for detectors necessitates substantial additional computational resources. Voxel-based detectors densely partition the point cloud in the height space and must use 3D convolutions. Therefore, compared with pillar-based 3D detectors, voxel-based 3D detectors generally achieve higher object detection accuracy, but their detection speed is much slower. Given these challenges, we propose a fully sparse ConvNeXt block for more efficient pillar feature extraction that selectively extracts features from effective data positions. We have developed SCN-Pillar, a pillar-based, fully sparse, lightweight 3D detector that adopts the sparse ConvNeXt. The SCN-Pillar has been validated on the Waymo open dataset, showcasing enhancements in accuracy across a range of object detection tasks. The APH improvement in pedestrian detection has reached more than 1.2. It only requires the computational cost of the pillar-based solution, yet its object detection accuracy exceeds that of the voxel-based solution. The object detection speed reaches 18.28 FPS. The code is available at https://github.com/kaikailab/SCN-Pillar.

The advent of 3D Gaussian Splatting (3DGS) has achieved a major breakthrough in the field of 3D reconstruction and realized real-time high-quality novel view synthesis. However, 3DGS relies on the color backpropagation gradients, which are greatly affected by the illumination intensity, to optimize the position parameters of gaussian spheres such as the average center coordinates and opacity, resulting in issues like blurring and artifacts when handling scenes with uneven illumination. This is particularly important in 3D reconstruction of autonomous driving scenarios, especially in night scenes with streetlights and vehicle lights. To address this issue, we propose the POGS framework, which integrates depth and edge constraints to optimize the position parameters of gaussian spheres and reduce the interference of illumination intensity. Our approach first incorporates a pre-trained monocular depth estimation network to generate depth maps, which are used to constrain the positions of gaussian spheres. In addition, we introduce an Edge Optimization Loss and incorporate the Segment Anything Model (SAM) to generate contour maps for constraining the weights of gaussian spheres, enabling the model to pay more attention to the gaussian spheres of the objects composing the scene. Our method significantly enhances the reconstruction performance of 3DGS on datasets with low texture, high depth of field, and low illumination, improving the SSIM by 3.1%, and outperforms the state-of-the-art method RawNeRF (2% SSIM) in NeRF while maintaining real-time rendering speed.

Image compression has been applied in the fields of image storage and video broadcasting. However, it’s formidably tough to distinguish the subtle quality differences between those distorted images generated by different algorithms. In this paper, we propose a new image quality assessment framework called Progressive Feature Fusion Network to decide which image is better in an image group. The network can effectively select clearer images based on their features. Experimental results show that compared with the current mainstream image quality assessment methods, the proposed network can achieve more accurate image quality assessment and ranks second in the benchmark of Challenge on Learned Image Compression in the image perceptual model track.

Deep learning with differential privacy (DP) has garnered significant attention over the past years, leading to the development of numerous methods aimed at enhancing model accuracy and training efficiency. This paper delves into the problem of training Transformer models with differential privacy. Our treatment is modular: the logic is to 'reduce' the problem of training DP Transformer to the more basic problem of training DP vanilla neural nets. The latter is better understood and amenable to many model-agnostic methods. Such 'reduction' is done by first identifying the hardness unique to DP Transformer training: the attention distraction phenomenon and a lack of compatibility with existing techniques for efficient gradient clipping. To deal with these two issues, we propose the Re-Attention Mechanism and Phantom Clipping, respectively. We believe that our work not only casts new light on training DP Transformers but also promotes a modular treatment to advance research in the field of differentially private deep learning.

With the rapid development of social networks, the harm caused by the leakage of personal sensitive information is becoming increasingly serious.In order to detect and identify personal sensitive information, existing methods build matching rules to detect specific sensitive entities and use machine learning methods to classify sensitive text.These methods face challenges in context analysis and adapting to Chinese language characteristics.This paper proposes CPSID, a method for detecting Chinese personal sensitive information.On the one hand, CPSID utilises rule matching to detect specific personal sensitive information only containing letters and numbers.More importantly, CPSID constructs a sequence labelling model named EBC (ELECTRA-BiLSTM-CRF) to detect more complex personal sensitive information that consist of Chinese characters.The EBC model uses the latest ELECTRA algorithm to implement word embedding, and uses BiLSTM and CRF models to extract personal sensitive information, which can detect Chinese sensitive entities accurately by analysing context information.The model achieves an F1 score of 94.09% on Chinese datasets, outperforming other similar models.Additionally, experiments on real data show CPSID has a better detection result than individual methods (rule matching or sequence labelling).

Network packets record communication behaviors and details, which is important for security audits, attack detection, and forensic analysis. For the effectiveness and timeliness of security analysis, it is necessary to fully store network packets and build an efficient packet index. However, the existing packet indexing algorithms based on the radix tree ignore the distribution characteristics of network traffic and use internal nodes with the same capacity for index construction, resulting in wasted disk space and poor retrieval performance. As a solution,$w$e propose ANTI, an adaptive network traffic indexing algorithm similar to Adaptive Radix Tree, which can adaptively switch internal nodes with different capacity according to the density of network traffic and compress the common prefix and distinct suffix of traffic attributes to balance the index construction performance and space utilization. We also implement a packet-aware network traffic archiving and indexing system to achieve full packet archival, efficient indexing, and fast retrieval. Finally, we empirically evaluate ANTI in IPv4 (IPv6) traffic scenarios, and the results confirm the effectiveness of ANTI as well as the benefit of adopting ANTI for enhancing indexing and retrieval performance compared with other state-of-art algorithms.

The extensive data collection performed by the Internet of Things (IoT) devices can put users at risk of data leakage. Consequently, IoT vendors are legally obliged to provide privacy policies to declare the scope and purpose of the data collection. However, complex and lengthy privacy policies are unfriendly to users, and the lack of a machine-readable format makes it difficult to check policy compliance automatically. To solve these problems, we first put forward a purpose-aware rule to formalize the purpose-driven data collection or use statement. Then, a novel approach to identify the rule from natural language privacy policies is proposed. To address the issue of diversity of purpose expression, we present the concepts of explicit and implicit purpose, which enable using the syntactic and semantic analyses to extract purposes in different sentences. Finally, the domain adaption method is applied to the semantic role labeling (SRL) model to improve the efficiency of purpose extraction. The experiments that are conducted on the manually annotated dataset demonstrate that this approach can extract purpose-aware rules from the privacy policies with a high recall rate of 91%. The implicit purpose extraction of the adapted model significantly improves the F1-score by 11%.

In order to improve the accuracy and efficiency of Android malware detection, an Android malware detection model based on decision tree (DT) with support vector machine (SVM) algorithm (DT-SVM) is proposed. Firstly, the original opcode, Dalvik opcode, is extracted by reversing Android software, and the eigenvector of the sample is generated by using the n-gram model. Then, a decision tree is generated via training the sample and updating decision nodes as SVM nodes from the bottom up according to the evaluation result of the test set in the decision path. The model effectively combines DT with SVM. Under the premise of maintaining a high-accuracy decision path, SVM is used to effectively reduce the overfitting problem in DT and thus improve the generalization ability, and maintain the superiority of SVM for the small sample training set. Finally, to test our approach, several simulation experiments are carried out, and the results demonstrate that the improved algorithm has better accuracy and higher speed as compared with other malware detection approaches.