大模型音频理解算法工程师-抖音音乐
Core
Develop large-scale automated audio quality monitoring and evaluation systems for music products using large language models and audio foundation models.
Role type
Senior IC machine-learning engineer (audio quality & large models)
Builds
Automated audio quality monitoring dashboards, defect detection systems, and quality assessment models for music products.
Domain
Music streaming / Audio AI / Large Language Models
Deliverable
production ML models
Required skills
Deep learning, Transformer architectures, Audio signal processing, Python, C/C++, PyTorch, Audio Foundation Models, Multi-modal large models, Large Language Models, Audio-Text Alignment, Reinforcement Learning, SFT, Preference Learning
Preferred skills
Audio Quality Assessment (AQA), MOS prediction, Audio Captioning, Audio Reasoning, Neural Codec, Audio enhancement, Audio generation, Research publications in top conferences
Responsibilities
Research and develop audio understanding algorithms for noise, distortion, and codec damage detection; Build scalable automated systems for quality trend analysis and problem clustering; Explore applications of large models in perceptual quality modeling and explainable audio analysis; Track and integrate international frontier research in audio foundation models.