光学工程硕士在读,方向明确,项目可查。 Optical engineering graduate student with focused, verifiable project work.
华侨大学光学工程硕士在读。方向包括图像恢复、光学设计和 AI 应用开发。这里放三类岗位的简历入口和项目证明。 I am pursuing a master’s degree in Optical Engineering at Huaqiao University. My work covers image restoration, optical design, and AI application development. This site organizes my resume and project evidence for three role directions.
三类岗位,三个入口。 Three role directions, three entry points.
首页用于快速判断。详细项目、排序和定向简历在各方向页面。 The homepage is for quick screening. Project details, ordering, and role-specific resumes are on each direction page.
技术支持方向 Technical Support Track
懂成像和光学,也做过现场支持。 Imaging and optics background, with on-site support experience.
查看定向页面 Open track page 岗位入口 Entry Route机器视觉方向 Machine Vision Track
光学机理、实验设计、模型指标都做过。 Hands-on work across optical mechanisms, experiments, and model metrics.
查看定向页面 Open track page 岗位入口 Entry Route产品 / AI 应用方向 Product / AI Applications Track
拆需求、做原型、做小程序,上线过可用版本。 Breaks down requirements, prototypes, and ships usable mini program versions.
查看定向页面 Open track page先看三项结果。 Start with three results.
论文指标、AI 小程序和投洽会经历分别对应技术、开发和现场支持。 Research metrics, the mini program, and expo service cover technical work, development, and on-site support.
基于深度学习的单镜头退化图像恢复研究 Single-Lens Degraded Image Deblurring via Deep Learning
研究单镜头成像中的空间非均匀模糊和高频衰减。完成退化分析、数据构建、模型训练、指标评估和论文投稿。 Studied spatially variant blur and high-frequency loss in single-lens imaging. Completed degradation analysis, dataset creation, model training, evaluation, and paper submission.
- 用 Python 与 PyTorch 搭建两阶段去模糊流程,构建约 4800 组 blur-sharp 成对数据。 Built a two-stage deblurring pipeline in Python and PyTorch with roughly 4,800 blur-sharp pairs.
- 设计无归一化残差主干,引入多尺度重建损失、边缘一致性与低权重对抗学习。 Designed a non-normalized residual backbone with multi-scale reconstruction loss, edge consistency, and low-weight adversarial learning.
- 完成消融实验、指标评估与论文撰写,并以第一作者发表于 IEEE 国际会议。 Completed ablation studies, evaluation, and paper writing, then published the work as first author at an IEEE conference.
知匣 AI 记忆盒子 Zhixia AI Memory Box Mini Program
针对 AI 对话上下文容易丢失的问题,独立做了一个微信小程序。功能包括存储、检索和管理 AI 会话内容。 Built a WeChat mini program for lost AI conversation context. It stores, retrieves, and manages AI-generated conversation records.
- 接入云数据库,实现标签分类、关键词搜索、置顶、回收站、导入导出与云同步。 Integrated cloud data storage with tagging, keyword search, pinning, recycle bin, import/export, and sync.
- 借助 GPT、Gemini、Codex 辅助需求拆解、代码生成、调试优化与快速迭代。 Used GPT, Gemini, and Codex to support requirement breakdown, code generation, debugging, and fast iteration.
- 7 天内完成可用版本上线,覆盖从问题发现到发布。 Shipped a usable version in seven days, from problem discovery to release.
投洽会国际嘉宾接待与现场支持 China International Fair Support & Guest Service
连续参加两届中国国际投资贸易洽谈会志愿服务。负责国际嘉宾接待、会务引导、签到支持和现场协调。 Served at two editions of the China International Fair for Investment and Trade. Responsibilities included guest reception, venue guidance, check-in support, and on-site coordination.
- 参与国际嘉宾接待与会务保障,曾服务匈牙利嘉宾。 Supported international guest reception and event operations, including Hungarian guests.
- 在多任务场景中维持秩序、信息传递和流程衔接。 Maintained order, information flow, and process continuity in multi-task environments.
- 负责嘉宾接待、签到支持和现场协调。 Handled guest reception, check-in support, and on-site coordination.
奖学金、优秀学生和志愿服务证明。 Scholarship, outstanding student, and volunteer service records.
这些材料用于补充学习成绩、执行稳定性和现场服务经历。 These records support academic performance, reliable execution, and on-site service experience.
研究生一等奖学业奖学金 First-Class Graduate Scholarship
前 10% Top 10%
优秀学生 Outstanding Student
前 5% Top 5%
金钥匙一星志愿者 Golden Key Volunteer
投洽会志愿服务荣誉 Recognition for fair volunteer service
下载对应简历。 Download the matching resume.
按岗位方向选择 PDF。需要沟通时,可直接发邮件。 Choose the PDF for the role direction. Email me if you want to discuss further.