Federated learning (FL) prevents direct exposure of raw data. However, it remains vulnerable to privacy and security threats such as inference and poisoning attacks. Traditional differential privacy (DP) methods utilize noise injection to mitigate these attacks, which inherently degrades the accuracy of the model. In this paper, we propose a robust FL framework with two alternative effective defense mechanisms to enhance privacy preservation for various scenarios. We first propose a dual-layer client-server collaborative differential privacy (CLDP). Clients utilize adaptive local differential privacy (LDP) for data privacy, while the server uses central differential privacy (CDP) on the global model to mitigate poisoning attacks. Second, we propose enhanced central differential privacy (ECDP), a layer-specific protection mechanism that strategically injects targeted noise into non-batch normalization layers to further preserve data privacy. To mitigate noise-induced model performance degradation, our solution combines personalized data transformation and gradient sparsification, effectively alleviating both non-IID data distribution skew and cumulative noise effects. Architecturally, we decentralize the federated learning system through edge node integration, thereby eradicating single points of failure. Experimental results demonstrate that our framework achieves a superior accuracy-privacy trade-off under strict privacy constraints, providing robust protection without compromising practical utility.

As a supervised learning algorithm, Support Vector Machine (SVM) is very popularly used for classification. However, the traditional SVM is error-prone because of easy to fall into local optimal solution. To overcome the problem, a new SVM algorithm based on Relief algorithm and particle swarm optimization-genetic algorithm (Relief-PGS) is proposed for feature selection and data classification, where the penalty factor and kernel function of SVM and the extracted feature of Relief algorithm are encoded as the particles of particle swarm optimization-genetic algorithm (PSO-GA) and optimized by iteratively searching for optimal subset of features. To evaluate the quality of features, Relief algorithm is used to screen the feature set to reduce the irrelevant features and effectively select the feature subset from multiple attributes. The advantage of Relief-PGS algorithm is that it can optimize both feature subset selection and SVM parameters including the penalty factor and the kernel parameter simultaneously. Numerical experimental results indicated that the classification accuracy and efficiency of Relief-PGS are superior to those of other algorithms including traditional SVM, PSO-GA-SVM, Relief-SVM, ACO-SVM, etc.

Depression is a common psychiatric disorder worldwide. However, in China, a considerable number of patients with depression are not diagnosed, and most of them are not aware of their depression. Despite increasing efforts, the goal of automatic depression screening from behavioral indicators has not been achieved. A major limitation is the lack of available multimodal depression corpus in Chinese since linguistic knowledge is crucial in clinical practice. Therefore, we first carried out a comprehensive survey with psychiatrists from a renowned psychiatric hospital to identify key interview topics which are highly related to the diagnosis of depression. Then, a semi-structural interview study was conducted over a year with subjects who have undergone clinical diagnosis and professional assessment. After that, Visual, acoustic, and textual features were extracted and analyzed between the two groups, statistically significant differences were observed in all three modalities. Benchmark evaluations of both single modal and multimodal fusion methods of depression assessment were also performed. A multimodal transformer-based fusion approach achieved the best performance. Finally, the proposed Chinese Multimodal Depression Corpus (CMDC) was made publicly available after de-identification and annotation. Hopefully, the release of this corpus would promote the research progress and practical applications of automatic depression screening.

Urban scene modeling is a challenging task for the photogrammetry and computer vision community due to its large scale, structural complexity, and topological delicacy. This paper presents an efficient multistep modeling framework for large-scale urban scenes from aerial images. It takes aerial images and a textured 3D mesh model generated by an image-based modeling system as the input and outputs compact polygon models with semantics at different levels of detail (LODs). Based on the key observation that urban buildings usually have piecewise planar rooftops and vertical walls, we propose a segment-based modeling method, which consists of three major stages: scene segmentation, roof contour extraction, and building modeling. By combining the deep neural network predictions with geometric constraints of the 3D mesh, the scene is first segmented into three classes. Then, for each building mesh, the 2D line segments are detected and used to slice the ground into polygon cells, followed by assigning each cell a roof plane via a MRF optimization. Finally, the LOD model is obtained by extruding cells to their corresponding planes. Compared with direct modeling in 3D space, we transform the mesh into a uniform 2D image grid representation and most of the modeling work is performed in 2D space, which has the advantages of low computational complexity and high robustness. In addition, our method doesn't require any global prior, such as the Manhattan or Atlanta world assumption, making it flexible to model scenes with different characteristics and complexity. Experiments on both single buildings and large-scale urban scenes demonstrate that by combining 2D photometric with 3D geometric information, the proposed algorithm is robust and efficient in urban scene LOD vectorized modeling compared with the state-of-the-art approaches.

Image based 3D modeling is an effective way to reconstruct large-scale scenes, especially city-level scenarios. In the image based modeling pipeline, obtaining a watertight mesh model from a noisy multiple view stereo point cloud is a key step to ensure the model quality. However, state-of-the-art method [34] relies on the global Delaunay-based optimization formed by all points and cameras, and will encounter scale problem when dealing with large scenes. To circumvent this limitation, this paper proposes a distributed surface reconstruction approach which could handle city-scale scenes with limited memory and time consumption. Firstly, the whole scene is adaptively divided into several chunks with overlapping boundaries, and each chunk can satisfy the memory limit. Then, the Delaunay-based optimization is performed to extract meshes for each chunk in parallel. Finally, the local meshes are merged together by resolving local inconsistencies in the overlapping areas. We test the proposed method on three city-scale scenes with billions of points and tens of thousands of images, and demonstrate its scalability and completeness compared with the state-of-the-art methods.