2027AIDU-多模态算法工程师(J99956)
Core
Research and iterate multimodal large models covering text-image, video, audio, and 3D fusion for understanding and generation.
Role type
Senior IC multimodal algorithm engineer (AIGC/Generative AI)
Builds
Multimodal models deployed in search, recommendation, AIGC, health, autonomous driving, and video understanding scenarios.
Domain
Artificial Intelligence / Computer Vision / Generative AI
Deliverable
production ML models
Required skills
Computer vision, image processing, deep learning, diffusion models, multimodal large models, model compression, Python, PyTorch, PaddlePaddle, TensorFlow, paper reproduction
Preferred skills
PhD in CS/AI, publications in top conferences (CVPR, NeurIPS, etc.), open-source contributions, experience in video generation or 3D generation
Technologies
CLIP, Flamingo, Qwen-VL, PyTorch, PaddlePaddle, TensorFlow
Responsibilities
Research cross-modal alignment, contrastive learning, and style transfer; build multimodal data pipelines and evaluation systems; optimize model training and inference efficiency; deploy models in core business scenarios.
Seniority
Senior, hands-on IC
