AIGC内容检测算法工程师（图像与视频方向）
Core
Develop and optimize AI-generated content (AIGC) detection models for images and videos to support content moderation, compliance labeling, and data filtering.
Role type
Senior IC machine-learning engineer (AIGC detection)
Builds
Production AIGC detection models and evaluation benchmarks
Domain
Content safety, image/video forensics, generative AI
Deliverable
production ML models
Required skills
AIGC detection model development, digital watermarking, statistical feature detection, image/video forensics (frequency domain analysis, noise residuals, GAN/Diffusion fingerprints), LLM/VLM fine-tuning and evaluation, data synthesis and pseudo-labeling, PyTorch, vLLM
Preferred skills
Multi-modal alignment and cross-modal detection, adversarial scenario optimization (local editing, compression, cropping/splicing), model training and inference strategy optimization
Technologies
PyTorch, vLLM, GAN, Diffusion
Responsibilities
Research and optimize AIGC detection models for images and videos; optimize model architecture, training, and inference strategies; handle complex adversarial scenarios including local editing and compression artifacts; establish AIGC detection evaluation systems and benchmarks.
Seniority
Senior, hands-on IC
