可控人像视频生成算法工程师（J91590）
Core
Develop and optimize algorithms for controllable portrait video generation, multi-modal driven video editing, and model training/inference within Baidu's AIGC and multi-modal systems.
Role type
Senior IC machine-learning engineer (video generation & editing)
Builds
Multi-modal portrait video generation products and video subject editing features
Domain
Computer Vision, AIGC, Multi-modal systems
Deliverable
production ML models
Required skills
Python, PyTorch, PaddlePaddle, TensorFlow, MXNet, image/video understanding, data production pipelines, model training, inference optimization
Preferred skills
Publications in top-tier conferences (CVPR, ICCV, ECCV, SIGGRAPH, TOG, TPAMI, ACL, EMNLP, NeurIPS), high rankings in coding/academic competitions (ACM, Kaggle)
Technologies
Python, PyTorch, PaddlePaddle, TensorFlow, MXNet
Responsibilities
Research and develop AIGC and multi-modal technologies; optimize model training and inference; support data production and quality improvement for video generation and editing products
Seniority
Senior, hands-on IC
