AIGC数字人视频生成算法实习生（J105810）
Core
Research and iterate on large models for generating digital human videos specifically for e-commerce scenarios.
Role type
Research intern (AIGC video generation)
Builds
Digital human video generation models for e-commerce
Domain
E-commerce, Generative AI, Video Generation
Deliverable
production ML models
Required skills
Deep learning, PyTorch, Diffusion Models, Video/Image generation technologies, Model inference acceleration, Academic research
Preferred skills
Experience with Wan/CogVideoX/Sora, Top-tier conference publications (CVPR/ICCV/NeurIPS/ECCV)
Technologies
PyTorch, Wan, H3, CogVideoX, Sora, IP-Adapter
Responsibilities
Research and implement inference acceleration strategies like DMD and sparse attention; Explore reference image control techniques; Develop super-resolution and local detail refinement models; Build training data pipelines and test sets; Track and write papers on frontier video generation topics.
