Xinjiang Institute of Engineering

ABSTRACT Phthalonitrile (PN)‐based fiber‐reinforced polymer composites (FRPs) have long been celebrated for exceptional thermal and mechanical performance, yet their practical deployment is hindered by persistent challenges in interfacial adhesion. Present research remains fragmented into disparate, case‐specific studies, leaving the modification‐strategy landscape and its intrinsic interconnections largely unexplored. This review therefore dissects the latest modification paradigms of FRPs, mapping how PN monomer design, matrix additives and fiber surface engineering translate from molecular‐scale innovations to macroscopic property gains. It first summarizes the effects of different PN monomers on the thermal and mechanical behavior of FRPs. A systematic taxonomy is then provided, contrasting matrix‐centric approaches—which elucidate modifier mechanisms and their impact on composite properties—with fiber‐centric techniques that enhance interfacial adhesion and interactions through specific surface treatments. Finally, the review outlines future research avenues that promise to push PN‐based FRPs beyond current performance ceilings, offering a roadmap for researchers and industries seeking next‐generation high‐performance materials.

DINOv3 has demonstrated remarkable capabilities in general object detection, its direct application to sheep face recognition faces two primary obstacles: the significant domain gap between general pre-training data and specific livestock scenarios, and the insufficiency in capturing fine-grained features required for distinguishing highly similar individuals. To address these challenges, this paper proposes an enhanced framework named AN-DINOv3. First, a feature pyramid module is introduced to capture multi-scale features, thereby significantly enhancing the model's ability to extract fine-grained details of sheep faces. Second, to bridge the domain gap, a LoRA strategy is incorporated to efficiently fine-tune the model to focus on the discriminative facial regions of sheep by mitigating domain drift problem. Extensive experiments on a self-built few-shot dataset demonstrate the superiority of the proposed AN-DINOv3method, achieving an accuracy of 94.5% and significantly outperforming mainstream baseline methods.