In response to the global energy crisis, high-performance transportation sectors are rapidly embracing lightweight materials to enhance energy efficiency and sustainability, while grappling with the persistent challenges of developing structural materials that meet stringent safety standards with robust mechanical performance and ease of scalability. Thus, this work presents a combined experimental and theoretical framework to develop a profound understanding of the synergistic effect in hybrid composites with bio-inspired fibrous reinforcements, by elucidating the interfacial interactions across multiple length-scales, encompassing atomic covalent bonding to micro-morphology. A model hybrid composite system, containing a self-assembled fibrous reinforcement consisting of nano-sized Graphene Nanoplatelets (GnP) covalently bonded onto chemically-modified micro-sized Glass Fibers (GF), was utilized to showcase the synergistic effect and highlight its associated mechanisms. The interfacial interactions of the reinforcement were optimized by obtaining the maximum density of covalent bonds, which was achieved with 0.5 wt% GnP for the hybrid composites containing 10 wt% GF, increasing the work of adhesion by 33 %, compared to the biphasic GF composites. The composite’s morphology contains minimal agglomeration with ∼68 % of GnPs oriented with the melt flow, supressing high-stress concentration areas, while the formed crystalline microstructure, with ∼18 % β-crystals, allows the matrix to absorb substantial energy. Furthermore, the increased trans-crystallization encapsulating the hierarchical reinforcement induced nanoscale stiffness variations, increasing rigidity, and forming an ∼16 µm gradient interphase that facilitates load transfer. The greatest synergistic effect observed was ∼54 %, ∼37 %, and ∼75 % for the tensile modulus, tensile strength, and impact strength, respectively. Additionally, a theoretical framework accounting for the synergistic effect was formulated, using a two-step core/shell homogenization model, which shows great potential in expediting the design and optimization of innovative hybrid composite materials.

Vision-language models (VLMs) often struggle to generate accurate and detailed captions for high-resolution images since they are typically pre-trained on low-resolution inputs (e.g., 224x224 or 336x336 pixels). Downscaling high-resolution images to these dimensions may result in the loss of visual details and the omission of important objects. To address this limitation, we propose a novel pipeline that integrates vision-language models, large language models (LLMs), and object detection systems to enhance caption quality. Our proposed pipeline refines captions through a novel, multi-stage process. Given a high-resolution image, an initial caption is first generated using a VLM, and key objects in the image are then identified by an LLM. The LLM predicts additional objects likely to co-occur with the identified key objects, and these predictions are verified by object detection systems. Newly detected objects not mentioned in the initial caption undergo focused, region-specific captioning to ensure they are incorporated. This process enriches caption detail while reducing hallucinations by removing references to undetected objects. We evaluate the enhanced captions using pairwise comparison and quantitative scoring from large multimodal models, along with a benchmark for hallucination detection. Experiments on a curated dataset of high-resolution images demonstrate that our pipeline produces more detailed and reliable image captions while effectively minimizing hallucinations.