arXiv subjects

Jingyi Yuan

Publications and source records attributed to Jingyi Yuan.

6 recordsLinked to original sources

arXiv2026

ReUnit: Multi-Granularity Visual Unitization for Long Video Understanding

Long-video understanding is constrained by the limited visual input capacity of video multimodal large language models (Video-MLLMs). Existing methods mainly optimize which content to retain, while the presentation of retained content often remains fixed. As a result, the same balance between spatial detail and content coverage is imposed across the entire visual input. We propose ReUnit, a training-free and query-aware framework that jointly determines which content to retain and how it should be presented. Guided by frame-level query relevance, ReUnit constructs and allocates visual units at multiple presentation granularities. More relevant content receives finer presentation, while broader context is represented more compactly. It realizes these granularities using visual units that carry one, four, or nine source frames and are each rendered as a standard image. Across four benchmarks and visual input budgets from 4 to 32 units, ReUnit achieves the highest average score at every tested budget. It improves over Uniform Sampling by 8.3--10.1 points on average, and the gains transfer across three additional Video-MLLM families. Project resources are available at https://github.com/charon525/ReUnit.

cs.CV↗

arXiv2025

AD-DINOv3: Enhancing DINOv3 for Zero-Shot Anomaly Detection with Anomaly-Aware Calibration

Zero-Shot Anomaly Detection (ZSAD) seeks to identify anomalies from arbitrary novel categories, offering a scalable and annotation-efficient solution. Traditionally, most ZSAD works have been based on the CLIP model, which performs anomaly detection by calculating the similarity between visual and text embeddings. Recently, vision foundation models such as DINOv3 have demonstrated strong transferable representation capabilities. In this work, we are the first to adapt DINOv3 for ZSAD. However, this adaptation presents two key challenges: (i) the domain bias between large-scale pretraining data and anomaly detection tasks leads to feature misalignment; and (ii) the inherent bias toward global semantics in pretrained representations often leads to subtle anomalies being misinterpreted as part of the normal foreground objects, rather than being distinguished as abnormal regions. To overcome these challenges, we introduce AD-DINOv3, a novel vision-language multimodal framework designed for ZSAD. Specifically, we formulate anomaly detection as a multimodal contrastive learning problem, where DINOv3 is employed as the visual backbone to extract patch tokens and a CLS token, and the CLIP text encoder provides embeddings for both normal and abnormal prompts. To bridge the domain gap, lightweight adapters are introduced in both modalities, enabling their representations to be recalibrated for the anomaly detection task. Beyond this baseline alignment, we further design an Anomaly-Aware Calibration Module (AACM), which explicitly guides the CLS token to attend to anomalous regions rather than generic foreground semantics, thereby enhancing discriminability. Extensive experiments on eight industrial and medical benchmarks demonstrate that AD-DINOv3 consistently matches or surpasses state-of-the-art methods.The code will be available at https://github.com/Kaisor-Yuan/AD-DINOv3.

cs.CV↗

arXiv2025

AFR-CLIP: Enhancing Zero-Shot Industrial Anomaly Detection with Stateless-to-Stateful Anomaly Feature Rectification

Recently, zero-shot anomaly detection (ZSAD) has emerged as a pivotal paradigm for industrial inspection and medical diagnostics, detecting defects in novel objects without requiring any target-dataset samples during training. Existing CLIP-based ZSAD methods generate anomaly maps by measuring the cosine similarity between visual and textual features. However, CLIP's alignment with object categories instead of their anomalous states limits its effectiveness for anomaly detection. To address this limitation, we propose AFR-CLIP, a CLIP-based anomaly feature rectification framework. AFR-CLIP first performs image-guided textual rectification, embedding the implicit defect information from the image into a stateless prompt that describes the object category without indicating any anomalous state. The enriched textual embeddings are then compared with two pre-defined stateful (normal or abnormal) embeddings, and their text-on-text similarity yields the anomaly map that highlights defective regions. To further enhance perception to multi-scale features and complex anomalies, we introduce self prompting (SP) and multi-patch feature aggregation (MPFA) modules. Extensive experiments are conducted on eleven anomaly detection benchmarks across industrial and medical domains, demonstrating AFR-CLIP's superiority in ZSAD.

cs.CV↗

arXiv2025

A Unified Approach to Enforce Non-Negativity Constraint in Neural Network Approximation for Optimal Voltage Regulation

Power system voltage regulation is crucial to maintain power quality while integrating intermittent renewable resources in distribution grids. However, the system model on the grid edge is often unknown, making it difficult to model physical equations for optimal control. Therefore, previous work proposes structured data-driven methods like input convex neural networks (ICNN) for "optimal" control without relying on a physical model. While ICNNs offer theoretical guarantees based on restrictive assumptions of non-negative neural network parameters, can one improve the approximation power with an extra step on negative duplication of inputs? We show that such added mirroring step fails to improve accuracy, as a linear combination of the original input and duplicated input is equivalent to a linear operation of ICNN's input without duplication. While this design can not improve performance, we propose a unified approach to embed the non-negativity constraint as a regularized optimization of the neural network, contrary to the existing methods, which added a loosely integrated second step for post-processing on parameter negation. Our integration directly ties back-propagation to simultaneously minimizing the approximation error while enforcing the convexity constraints. Numerical experiments validate the issues of the mirroring method and show that our integrated objective can avoid problems such as unstable training and non-convergence existing in other methods for optimal control. (preprint)

eess.SY↗

arXiv2024

A ground-based dataset and a diffusion model for on-orbit low-light image enhancement

On-orbit service is important for maintaining the sustainability of space environment. Space-based visible camera is an economical and lightweight sensor for situation awareness during on-orbit service. However, it can be easily affected by the low illumination environment. Recently, deep learning has achieved remarkable success in image enhancement of natural images, but seldom applied in space due to the data bottleneck. In this article, we first propose a dataset of the Beidou Navigation Satellite for on-orbit low-light image enhancement (LLIE). In the automatic data collection scheme, we focus on reducing domain gap and improving the diversity of the dataset. we collect hardware in-the-loop images based on a robotic simulation testbed imitating space lighting conditions. To evenly sample poses of different orientation and distance without collision, a collision-free working space and pose stratified sampling is proposed. Afterwards, a novel diffusion model is proposed. To enhance the image contrast without over-exposure and blurring details, we design a fused attention to highlight the structure and dark region. Finally, we compare our method with previous methods using our dataset, which indicates that our method has a better capacity in on-orbit LLIE.

cs.CV↗

arXiv2019

Quantifying Hosting Capacity for Rooftop PV System in LV Distribution Grids

Power systems face increasing challenges on reliable operations due to the widespread distributed generators (DGs), e.g., rooftop PV system in the low voltage (LV) distribution grids. Characterizing the hosting capacity (HC) is vital for assessing the total amount of distributed generations that a grid can hold before upgrading. For analyzing HC, some methods conduct extensive simulations, lacking theoretical guarantees and can time-consuming. Therefore, there are also methods employing optimization over all necessary operation constraints. But, the complexity and inherent non-convexity lead to non-optimal solutions. To solve these problems, this paper provides a constructive model for HC determination. Based on geometrically obtained globally optimal HC, we construct HC solutions sequentially according to realistic constraints, so that we can obtain optimal solution even with non-convex model. For practical adaption, we also consider three-phase unbalance condition and parallel computation to speed-up. We validated our method by extensive numerical results on IEEE standard systems, such as 8-bus and 123-bus radial distribution grids, where the global optimums are obtained and performance illustrations are demonstrated.

eess.SY↗