Recent advances in parameter-efficient transfer learning have demonstrated the utility of composing LoRA adapters from libraries of pretrained modules. However, most existing approaches rely on simple retrieval heuristics or uniform averaging, which overlook the latent structure of task relationships in representation space. We propose a new framework for adapter reuse that moves beyond retrieval, formulating adapter composition as a geometry-aware sparse reconstruction problem. Specifically, we represent each task by a latent prototype vector derived from the base model’s encoder and aim to approximate the target task prototype as a sparse linear combination of retrieved reference prototypes, under an L1-regularized optimization objective. The resulting combination weights are then used to blend the corresponding LoRA adapters, yielding a composite adapter tailored to the target task. This formulation not only preserves the local geometric structure of the task representation manifold, but also promotes interpretability and efficient reuse by selecting a minimal set of relevant adapters. We demonstrate the effectiveness of our approach across multiple domains—including medical image segmentation, medical report generation and image synthesis. Our results highlight the benefit of coupling retrieval with latent geometry-aware optimization for improved zero-shot generalization.

Abstract With the continuous integration of functions in wearable devices, power consumption demands have increased significantly, posing serious challenges to conventional power supply methods. Wearable self-powered technologies offer an effective solution to this issue. This study focuses on the efficient utilization of solar energy from the human wrist and presents the design and implementation of a flexible photovoltaic wristband. The wristband employs a multi-directional parallel array of photovoltaic cells, integrated with an energy management module, enabling it to adapt effectively to the dynamic and non-uniform solar irradiance conditions on the wrist. Through both simulated sunlight and real outdoor environment tests, the energy harvesting and load-driving performances of the photovoltaic wristband were comprehensively evaluated. The results show that under a highest average outdoor illuminance of 37.38 × 10 3 lx (525 W·m −2 ), the wristband delivers an average output power of 15.88 mW, providing a stable 3.3 V supply to a wearable motion sensing node, thereby enabling self-powered operation. During a complete “energy accumulation-load activation” cycle, the sensing node can operate for 37.84 s to perceive and transmit data. By employing a one-dimensional convolutional neural networks (1D-CNN) algorithm, accurate recognition of four motion states is successfully achieved. This work presents a systematic study covering energy harvesting scenarios analysis, wristband design, performance evaluation, and sensing application. The proposed photovoltaic wristband demonstrates excellent cyclic energy accumulation and stable power supply capabilities, validating the feasibility and practicality of the wearable photovoltaic self-powered system and highlighting its promising potential for future wearable applications.

Compared with botanical medicines, the intelligent identification of animal-derived medicines has developed relatively slowly and presents greater challenges due to species diversity, substantial morphological variations after processing, and the prevalence of adulteration. Tortoise plastron is a representative animal-derived medicine whose subtle morphological differences complicate reliable authentication. This study aimed to establish a standardized image-based framework to achieve accurate and reproducible identification of tortoise plastron and its common adulterants. An RGB image dataset covering Chinemys reevesii , Mauremys mutica , Ocadia sinensis , Malayemys subtrijuga , and Trachemys scripta elegans was constructed. Images were enhanced through geometric transformations, color jittering, and Gaussian blurring to simulate diverse acquisition conditions. Segmentation was performed using the SAM2 model to remove background noise and extract core regions, and classification was conducted using YOLO11 models of three scales (m, l, x). Training and validation were carried out on datasets with and without segmentation-based augmentation, each repeated three times, with average performance recorded. Statistical analysis included comparison of model performance metrics across datasets and model scales. The combined preprocessing–segmentation–classification workflow effectively captured both global and fine-grained features of tortoise plastron. YOLO11m with segmentation-based augmentation achieved the most balanced performance across accuracy and robustness. Eight technical modules, including attention-enhanced feature extraction and multi-scale pooling, contributed to improved classification precision. A practical recognition application was developed to facilitate user-friendly deployment. This study established a comprehensive digital framework for the objective identification of tortoise plastron and adulterants, transforming subjective trait-based evaluation into quantitative image analysis. The integration of advanced segmentation and multi-scale feature fusion provides a transferable paradigm for the intelligent identification of animal-derived medicines, with potential to enhance quality control and authenticity assurance in traditional Chinese medicine.

Echocardiography (ECHO) is essential for cardiac assessments, but its video quality and interpretation heavily relies on manual expertise, leading to inconsistent results from clinical and portable devices. ECHO video generation offers a solution by improving automated monitoring through synthetic data and generating high-quality videos from routine health data. However, existing models often face high computational costs, slow inference, and rely on complex conditional prompts that require experts' annotations. To address these challenges, we propose ECHOPulse, an ECG-conditioned ECHO video generation model. ECHOPulse introduces two key advancements: (1) it accelerates ECHO video generation by leveraging VQ-VAE tokenization and masked visual token modeling for fast decoding, and (2) it conditions on readily accessible ECG signals, which are highly coherent with ECHO videos, bypassing complex conditional prompts. To the best of our knowledge, this is the first work to use time-series prompts like ECG signals for ECHO video generation. ECHOPulse not only enables controllable synthetic ECHO data generation but also provides updated cardiac function information for disease monitoring and prediction beyond ECG alone. Evaluations on three public and private datasets demonstrate state-of-the-art performance in ECHO video generation across both qualitative and quantitative measures. Additionally, ECHOPulse can be easily generalized to other modality generation tasks, such as cardiac MRI, fMRI, and 3D CT generation. We will make the synthetic ECHO dataset, along with the code and model, publicly available upon acceptance.

Ensuring fairness in medical image segmentation is critical due to biases in imbalanced clinical data acquisition caused by demographic attributes (e.g., age, sex, race) and clinical factors (e.g., disease severity). To address these challenges, we introduce Distribution-aware Mixture of Experts (dMoE), inspired by optimal control theory. We provide a comprehensive analysis of its underlying mechanisms and clarify dMoE's role in adapting to heterogeneous distributions in medical image segmentation. Furthermore, we integrate dMoE into multiple network architectures, demonstrating its broad applicability across diverse medical image analysis tasks. By incorporating demographic and clinical factors, dMoE achieves state-of-the-art performance on two 2D benchmark datasets and a 3D in-house dataset. Our results highlight the effectiveness of dMoE in mitigating biases from imbalanced distributions, offering a promising approach to bridging control theory and medical image segmentation within fairness learning paradigms. The source code is available at https://github.com/tvseg/dMoE.

Solving inverse problems—recovering signals from incomplete or noisy measurements—is fundamental in science and engineering. Score-based generative models (SGMs) have recently emerged as a powerful framework for this task. Two main paradigms have formed: unsupervised approaches that adapt pretrained generative models to inverse problems, and supervised bridge methods that train stochastic processes conditioned on paired clean and corrupted data. While the former typically assume knowledge of the measurement model, the latter have largely overlooked this structural information. We introduce System-embedded Diffusion Bridge Models (SDBs), a new class of supervised bridge methods that explicitly embed the known linear measurement system into the coefficients of a matrix-valued SDE. This principled integration yields consistent improvements across diverse linear inverse problems and demonstrates robust generalization under system misspecification between training and deployment, offering a promising solution to real-world applications.

Diffusion models have achieved state-of-the-art performance in generative modeling, yet their sampling procedures remain vulnerable to hallucinations—often stemming from inaccuracies in score approximation. In this work, we reinterpret diffusion sampling through the lens of optimization and introduce RODS (Robust Optimization–inspired Diffusion Sampler), a novel method that detects and corrects high-risk sampling steps using geometric cues from the loss landscape. RODS enforces smoother sampling trajectories and \textit{adaptively} adjusts perturbations, reducing hallucinations without retraining and at minimal additional inference cost. Experiments on AFHQv2, FFHQ, and 11k-hands demonstrate that RODS maintains comparable image quality and preserves generation diversity. More importantly, it improves both sampling fidelity and robustness, detecting over 70\% of hallucinated samples and correcting more than 25\%, all while avoiding the introduction of new artifacts. We release our code at https://github.com/Yiqi-Verna-Tian/RODS.

Despite achieving impressive results in general-purpose semantic segmentation with strong generalization on natural images, the Segment Anything Model (SAM) has shown less precision and stability in medical image segmentation. In particular, the original SAM architecture is designed for 2D natural images and is therefore not support to handle three-dimensional information, which is particularly important for medical imaging modalities that are often volumetric or video data. In this paper, we introduce MediViSTA, a parameter-efficient fine-tuning method designed to adapt the vision foundation model for medical video, with a specific focus on echocardiography segmentation. To achieve spatial adaptation, we propose a frequency feature fusion technique that injects spatial frequency information from a CNN branch. For temporal adaptation, we integrate temporal adapters within the transformer blocks of the image encoder. Using a fine-tuning strategy, only a small subset of pre-trained parameters is updated, allowing efficient adaptation to echocardiography data. The effectiveness of our method has been comprehensively evaluated on three datasets, comprising two public datasets and one multi-center in-house dataset. Our method consistently outperforms various state-of-the-art approaches without using any prompts. Furthermore, our model exhibits strong generalization capabilities on unseen datasets, surpassing the second-best approach by 2.15% in Dice and 0.09 in temporal consistency. The results demonstrate the potential of MediViSTA to significantly advance echocardiography video segmentation, offering improved accuracy and robustness in cardiac assessment applications.

Echocardiography is the first-line non-invasive cardiac imaging modality, providing rich spatio-temporal information on cardiac anatomy and physiology. Recently, foundation model trained on extensive and diverse datasets has shown strong performance in various downstream tasks. However, translating foundation models into the medical imaging domain remains challenging due to domain differences between medical and natural images, the lack of diverse patient and disease datasets. In this paper, we introduce EchoFM, a general-purpose vision foundation model for echocardiography trained on a large-scale dataset of over 20 million echocardiographic images from 6,500 patients. To enable effective learning of rich spatio-temporal representations from periodic videos, we propose a novel self-supervised learning framework based on a masked autoencoder with a spatio-temporal consistent masking strategy and periodic-driven contrastive learning. The learned cardiac representations can be readily adapted and fine-tuned for a wide range of downstream tasks, serving as a strong and flexible backbone model. We validate EchoFM through experiments across key downstream tasks in the clinical echocardiography workflow, leveraging public and multi-center internal datasets. EchoFM consistently outperforms SOTA methods, demonstrating superior generalization capabilities and flexibility. The code and checkpoints are available at: https://github.com/SekeunKim/EchoFM.git.

The prevalence of location-based social networks gives rise to the study of Geo-Social Keyword Query (GSKQ), where the Reverse Top-k Geo-Social Keyword Query (RkGSKQ) is a key technique used to detect prospective customers. Existing RkGSKQ solutions only focus on query efficiency, but ignore the quality of query results. When the query issuer obtained unexpected query results, no suggestion was offered to aid them get better ones. Thus, the overall utility of this query remains a problem. Towards this end, this paper considers the usability of RkGSKQ results and study two novel problems, i.e., maximizing the size of RkGSKQ results and why-not questions on RkGSKQ, both of which have potential applications in market analysis.

Ridesharing refers to a transportation scenario where travellers with similar itineraries and time schedules share a vehicle for a trip and split the travel cost, which may include fuel, tolls, and parking fees. Ridesharing is popular among travellers because it can reduce their travel costs, and it also holds the potential to reduce travel time, congestion, air pollution, and overall fuel consumption. However, existing ridesharing systems often offer each traveller only one choice that aims to minimize system-wide vehicle travel distance or time. We propose a solution that offers more options. Specifically, we do this by considering both pick-up time and price, so that travellers are able to choose the vehicle that matches their preferences best. In order to identify quickly vehicles that satisfy incoming ridesharing requests, we propose two efficient matching algorithms that follow the single-side and dual-side search paradigms, respectively. To further accelerate the matching, indexes on the road network and vehicles are developed, based on which several pruning heuristics are designed. Extensive experiments on a large Shanghai taxi dataset offer insights into the performance of our proposed techniques and compare with a baseline that extends the state-of-the art method.