卢英豪，目前就读于宁波大学信息科学与工程学院，是计算机技术专业硕士研究生，在王翀老师的指导下开展计算机视觉研究。

研究方向聚焦人物交互检测（Human-Object Interaction Detection, HOI），主要围绕开放词汇场景下的人—物关系理解、视觉语言对齐和未见类别泛化展开。项目中以“人、物体、动作谓词”三元组为核心，结合 CLIP 等视觉语言模型建模视觉特征、文本语义、类别先验和交互关系之间的映射，关注长尾类别、零样本迁移和复杂场景下的关系判别能力。

实习和项目经历偏向工业视觉与视觉算法工程，参与过 AGV 多镜头货架图像拼接、一维码运动模糊复原、盖板玻璃缺陷检测和食堂餐盘营养分析等项目。工作内容覆盖图像采集与预处理、相机标定、特征匹配、目标检测、缺陷分类、图像复原、模型训练评估、误检漏检分析、结果可视化和 PyQt 工具开发，能够围绕真实数据搭建从算法验证、参数调试的完整流程。

Hello, I am Yinghao Lu, a Master student in Computer Technology at the Faculty of Electrical Engineering and Computer Science, Ningbo University, advised by Prof. Chong Wang.

My research focuses on Human-Object Interaction Detection (HOI), with an emphasis on human-object relationship understanding, vision-language alignment, and unseen-category generalization in open-vocabulary settings. Centered on the human-object-action triplet, I explore CLIP-based vision-language models to map visual features, textual semantics, category priors, and interaction relations, with attention to long-tail categories, zero-shot transfer, and relation reasoning in complex scenes.

My internships and projects are oriented toward industrial vision and vision algorithm engineering, including AGV multi-camera shelf stitching, motion-blurred barcode restoration, cover-glass defect detection, and cafeteria tray nutrition analysis. My work spans image acquisition and preprocessing, camera calibration, feature matching, object detection, defect classification, image restoration, model training and evaluation, false-positive and false-negative analysis, result visualization, and PyQt tool development, covering the workflow from algorithm validation and parameter tuning to prototype delivery.