One-shot Federated Learning (OFL) has emerged as a promising paradigm, enabling global model training with minimal communication overhead. In OFL, the server model is usually distilled from an ensemble of pre-trained client models, while the ensemble also facilitates synthetic data generation for the knowledge distillation process. Prior works show that the performance of the final model is fundamentally tied to both the quality of the synthetic data and the ensemble. However, existing methods often optimize these two components separately, overlooking their interaction. To address this coupled optimization problem and provide a unified solution to the dual challenges of data and model heterogeneity inherent in OFL, we introduce Co-Boosting++, a novel OFL framework where synthetic data generation and ensemble construction mutually enhance each other in an iterative fashion. First, we fix the ensemble and generate hard samples in an adversarial manner. These samples are crucial for enhancing the robustness of knowledge transfer, as they challenge the model to generalize better, thereby improving quality of the synthetic data and subsequent distillation process. Second, leveraging these hard samples, we enhance the ensemble via a Mixture of Experts (MoE) mechanism. MoE allows dynamic adjustment of ensemble weights based on the generated hard samples, which enables the ensemble to better capture diverse and heterogeneous knowledge from client models. Furthermore, we extend Co-Boosting++ to support the simultaneous generation of multiple heterogeneous target models, enabling efficient adaptation to diverse device constraints. Extensive experiments on benchmark datasets demonstrate that Co-Boosting++ consistently outperforms state-of-the-art methods due to its coupled optimization of data and ensemble quality. Additionally, Co-Boosting++ is highly practical in real-world model market scenarios, requiring no local training modifications, additional transmissions, or restrictions on client model architectures.

This article introduces a multiview 3-D microscopic profilometry system equipped with tilted cameras adhering to the Scheimpflug condition and a vertically aligned projector. It utilizes a novel 3-D cross-ratio invariant (3D-CRI) model that offers inherent self-correction, enhanced global consistency, and computational efficiency. The inherent self-correction mitigates optical contaminations like multiple reflections by using deviations from the epipolar line for optimal candidate selection. The model is also designed as spatial lines to associate the approximate linear error distribution across depths, thereby facilitating its further correction (e.g., proposed embedded linear compensation method to address data inconsistency among various binocular projector-camera setups). Additionally, the model simplifies computational demands by directly correlating phase differences to spatial coordinates. Furthermore, a conversion method from the triangular stereo to the 3D-CRI model is developed, avoiding the complex per-pixel calibration. The experiments are conducted to verify robustness, globally consistent accuracy, and speed performance. The experimental results validate that the system is robust to multiple reflections by testing on the printed circuit boards with dense components. The global consistency is also verified by the experiments with better quantitative metrics within the whole measurement range over ten times the depth of field. In particular, the quantitative metrics of theZ-axis are halved. The speed performance shows that our method is the fastest among competing techniques.

Underwater object detection is critical for applications, such as environmental monitoring, resource exploration, and the navigation of autonomous underwater vehicles. However, accurately detecting small objects in underwater environments remains challenging due to noisy imaging conditions, variable illumination, and complex backgrounds. To address these challenges, we propose the adaptive residual attention network (ARAN), an optimized deep learning framework designed to enhance the detection and precise identification of diminutive targets in complex aquatic settings. ARAN incorporates the proposed Fusion path aggregation network (PANet), which refines spatial features by effectively distinguishing objects from their backgrounds. The framework integrates three novel modules: first, multiscale feature attention, which enhances low-level feature extraction; second, high–low feature residual learning, which rearranges channel and batch dimensions to capture pixel-level relationships through cross-dimensional interactions; and third, multilevel feature dynamic aggregation, which dynamically adjusts fusion weights to facilitate progressive multilevel feature fusion and mitigate conflicts in multiscale integration, ensuring that small objects are not overshadowed. Extensive experiments on four benchmark datasets demonstrate that ARAN significantly outperforms mainstream models, achieving state-of-the-art performance. Notably, on the CSIRO dataset, ARAN attains a mean average precision at 50% of 98%, precision of 94.7%, F2-score of 94.6%, and recall of 94.7%. These results confirm our model's superior accuracy, robustness, and efficiency in underwater object detection, highlighting its potential for practical deployment in challenging aquatic environments. We will release the code on GitHub upon acceptance of the article.

In 3D microscopic imaging, the extremely shallow depth of field presents a challenge for accurate 3D reconstruction in cases of significant defocus. Traditional calibration methods rely on the spatial extraction of feature points to establish spatial 3D information as the optimization objective. However, these methods suffer from reduced extraction accuracy under defocus conditions, which causes degradation of calibration performance. To extend calibration volume without compromising accuracy in defocused scenarios, we propose a per-pixel calibration based on multi-view 3D reconstruction errors. It utilizes 3D reconstruction errors among different binocular setups as an optimization objective. We first analyze multi-view 3D reconstruction error distributions under the poor-accuracy optical model by employing a multi-view microscopic 3D measurement system using telecentric lenses. Subsequently, the 3D proportion model is proposed for implementing our error-based per-pixel calibration, derived as a spatial linear expression directly correlated with the 3D reconstruction error distribution. The experimental results confirm the robust convergence of our method with multiple binocular setups. Near the focus volume, the multi-view 3D reconstruction error remains approximately m (less than 0.5 camera pixel pitch), with absolute accuracy maintained within 0.5% of the measurement range. Beyond tenfold depth of field, the multi-view 3D reconstruction error increases to around m (still less than 2 camera pixel pitches), while absolute accuracy remains within 1% of the measurement range. These high-precision measurement results validate the feasibility and accuracy of our proposed calibration.

One-shot Federated Learning (OFL) has become a promising learning paradigm, enabling the training of a global server model via a single communication round. In OFL, the server model is aggregated by distilling knowledge from all client models (the ensemble), which are also responsible for synthesizing samples for distillation. In this regard, advanced works show that the performance of the server model is intrinsically related to the quality of the synthesized data and the ensemble model. To promote OFL, we introduce a novel framework, Co-Boosting, in which synthesized data and the ensemble model mutually enhance each other progressively. Specifically, Co-Boosting leverages the current ensemble model to synthesize higher-quality samples in an adversarial attack manner. These hard samples are then employed to promote the quality of the ensemble model by adjusting the ensembling weights for each client model. Consequently, Co-Boosting periodically achieves high-quality data and ensemble models. Extensive experiments demonstrate that Co-Boosting can substantially outperform existing baselines under various settings. Moreover, Co-Boosting eliminates the need for adjustments to the client's local training, requires no additional data or model transmission, and allows client models to have heterogeneous architectures.

Federated learning (FL) is a promising framework for privacy-preserving and distributed training with decentralized clients. However, there exists a large divergence between the collected local updates and the expected global update, which is known as the client drift and mainly caused by heterogeneous data distribution among clients, multiple local training steps, and partial client participation training. Most existing works tackle this challenge based on the empirical risk minimization (ERM) rule, while less attention has been paid to the relationship between the global loss landscape and the generalization ability. In this work, we propose FedGAMMA, a novel FL algorithm with Global sharpness-Aware MiniMizAtion to seek a global flat landscape with high performance. Specifically, in contrast to FedSAM which only seeks the local flatness and still suffers from performance degradation when facing the client-drift issue, we adopt a local varieties control technique to better align each client's local updates to alleviate the client drift and make each client heading toward the global flatness together. Finally, extensive experiments demonstrate that FedGAMMA can substantially outperform several existing FL baselines on various datasets, and it can well address the client-drift issue and simultaneously seek a smoother and flatter global landscape.

Personalized federated learning is proposed to handle the data heterogeneity problem amongst clients by learning dedicated tailored local models for each user. However, existing works are often built in a centralized way, leading to high communication pressure and high vulnerability when a failure or an attack on the central server occurs. In this work, we propose a novel personalized federated learning framework in a decentralized (peer-to-peer) communication protocol named DisPFL, which employs personalized sparse masks to customize sparse local models on the edge. To further save the communication and computation cost, we propose a decentralized sparse training technique, which means that each local model in DisPFL only maintains a fixed number of active parameters throughout the whole local training and peer-to-peer communication process. Comprehensive experiments demonstrate that DisPFL significantly saves the communication bottleneck for the busiest node among all clients and, at the same time, achieves higher model accuracy with less computation cost and communication rounds. Furthermore, we demonstrate that our method can easily adapt to heterogeneous local clients with varying computation complexities and achieves better personalized performances.

Abstract Similar to language and music, dance performances provide an effective way to express human emotions. With the abundance of the motion capture data, content‐based motion retrieval and classification have been fiercely investigated. Although researchers attempt to interpret body language in terms of human emotions, the progress is limited by the scarce 3D motion database annotated with emotion labels. This article proposes a hybrid feature for emotional classification in dance performances. The hybrid feature is composed of an explicit feature and a deep feature. The explicit feature is calculated based on the Laban movement analysis, which considers the body, effort, shape, and space properties. The deep feature is obtained from latent representation through a 1D convolutional autoencoder. Eventually, we present an elaborate feature fusion network to attain the hybrid feature that is almost linearly separable. The abundant experiments demonstrate that our hybrid feature is superior to the separate features for the emotional classification in dance performances.

Batch measurement of sheet-like product quantity is very common in manufacturing and commercial areas. Particularly, to satisfy the growing requirement of automatic and nondestructive detection, a machine vision system for stacked substrates counting is proposed in this paper. With a brief description of the system architecture and imaging module, the challenges for real data analysis are investigated. Our main contribution is to develop a robust stripe detection algorithm by combining the merits of local template matching and global frequency domain filtering. Besides the peak-valley feature across the stack profile, the collinear prior shape along the substrates is also utilized in a morphological scheme to purify the obtained ridge-line or stripe images, which, finally, benefits the statistical counting output. It is verified in experiments using a diversity of substrate media that our developed system can achieve a high counting accuracy with the detection error not more than 0.01% as the sample thickness varies between 0.05 mm and 0.5 mm. Moreover, the proposed algorithm performs much robustly with various kinds of abnormalities and interferences like irregular piling, media distortion, and foreign contamination.