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.

MFP-CLIP: Exploring the Efficacy of Multi-Form Prompts for Zero-Shot Industrial Anomaly Detection

Recently, zero-shot anomaly detection (ZSAD) has emerged as a pivotal paradigm for identifying defects in unseen categories without requiring target samples in training phase. However, existing ZSAD methods struggle with the boundary of small and complex defects due to insufficient representations. Most of them use the single manually designed prompts, failing to work for diverse objects and anomalies. In this paper, we propose MFP-CLIP, a novel prompt-based CLIP framework which explores the efficacy of multi-form prompts for zero-shot industrial anomaly detection. We employ an image to text prompting(I2TP) mechanism to better represent the object in the image. MFP-CLIP enhances perception to multi-scale and complex anomalies by self prompting(SP) and a multi-patch feature aggregation(MPFA) module. To precisely localize defects, we introduce the mask prompting(MP) module to guide model to focus on potential anomaly regions. Extensive experiments are conducted on two wildly used industrial anomaly detection benchmarks, MVTecAD and VisA, demonstrating MFP-CLIP's superiority in ZSAD.

Diffusion Model Based Low-Light Image Enhancement for Space Satellite

Space-based visible camera is an important sensor for space situation awareness during proximity operations. However, visible camera can be easily affected by the low illumination in the space environment. Recently, deep learning approaches have achieved remarkable success in image enhancement of natural images datasets, but seldom applied in space due to the data bottleneck. In this article, we propose a data-driven method for low-light image enhancement (LLIE) of spin targets in space environment based on diffusion model. Firstly, a dataset collection scheme is devised. To reduce the domain gap and improve the diversity and quality of the dataset, we collect the data with the camera on a ground-test system imitating the low lighting conditions and relative attitude change of satellite in space. The satellite motion is controlled by a 6-DoF robot. To generate different poses, a advanced sampling method is combined with collision detection in physical simulation. The entire process is automated. Based on our dataset, a novel diffusion model is proposed. The diffusion and denoising process are directly conducted on the grayscale channel to save computational resources. To take advantage of the inner information of RGB channels, we rescale the RGB feature maps and insert them into the downsampling layers to help feature extraction. The enhanced results with our method have been verified to be better in image light enhancement and competitive in image quality compared with previous methods. To the best of our knowledge, this is the first work of LLIE using diffusion model.