高级视频算法工程师 - AIGC视频生成

负责多模态内容理解与视频智能混剪算法的研发和迭代，持续提升模型对视频语义、节奏和审美偏好的理解能力，驱动剪辑效果的智能化和个性化；

协同工程与产品团队，将算法方案从验证转化为可上线部署的服务能力，确保模型在生产环境中的性能、稳定性与推理效率满足业务要求；

跟踪并评估AIGC领域的前沿进展，针对产品中长期需求完成技术预研与可行性验证，为关键模块的架构演进提供方案支撑

计算机相关专业本科及以上学历，5年以上算法研发经验，有完整参与算法项目从设计到上线全流程的实践经历；

掌握Transformer、DiT、Diffusion、CNN等深度神经网络模型原理，具备实际项目应用经验，对机器学习与图像、视频处理算法有系统性理解；

熟练使用PyTorch或TensorFlow等深度学习框架，具备扎实的算法实现能力，熟悉模型压缩、量化、部署等性能调优常用手段，能独立承担复杂算法模块的开发与优化；

具备较强的分析与问题定位能力，能熟练阅读并理解领域内英文技术文献或论文，从中提炼可落地的技术思路；

具备良好的沟通与协作素养，能够在跨职能团队中高效协同，共同攻克技术难点。 加分项： · 有视频内容理解、视频生成、文生图或多模态大模型相关项目的研发与落地经验；

在CVPR、ICCV、ECCV等计算机视觉领域顶级会议上以第一作者或共同作者身份发表过论文。 Senior Video Algorithm Engineer – AIGC Role Overview: As a Senior Video Algorithm Engineer on the AIGC team, you will research and develop core algorithms in multi-modal content understanding and intelligent video creation, delivering technical solutions that power Meitu’s global imaging products. You’ll work at the intersection of computer vision, generative models, and large-scale deployment, helping shape how millions of users create and interact with visual content. Key Responsibilities: · Drive research, development, and continuous optimization of advanced algorithms for multi-modal content understanding and intelligent video editing within the AIGC project team;

Translate algorithmic advances into product-ready features, ensuring efficient deployment and measurable product impact. Requirements: · Bachelor’s degree or higher in Computer Science or a related field, with 5+ years of hands-on experience delivering algorithm projects from research to production deployment;

Comprehensive knowledge of deep neural network architectures such as Transformers, DIT, Diffusion models, and CNNs, with practical project application; deep understanding of machine learning and image/video processing algorithms;

Proficient in PyTorch or TensorFlow; strong algorithm implementation and performance optimization skills, able to independently design, develop, and fine-tune complex algorithm modules;

Strong analytical and problem-solving skills, with the ability to read and interpret English-language academic papers in the field;

Effective collaborator who communicates clearly and works efficiently across functions to tackle challenging technical problems. Preferred Qualifications: · Practical experience in video content understanding, video generation, text-to-image generation, multi-modal modeling, or related areas;

Publications in top-tier computer vision conferences (e.g., CVPR, ICCV, ECCV). Nice-to-Haves section was suggested by AI based on the JD content — please review before publishing.