Computer Vision ML Engineer

该职位来源于猎聘 **About the Role** We're looking for a Computer Vision Engineer to build and deploy machine vision systems for automated optical inspection (AOI) in high-precision manufacturing. You'll own the full pipeline — from imaging and data to models running in real time on the production line — and work directly with product managers, software, data and field engineers to deliver end-to-end solutions. **What You'll Do**

• Design, train, and deploy deep learning models for defect detection, classification, and segmentation on production imagery
• Conduct research and development in emerging AI technologies like vision language models and multi-modal language models
• Build classical + ML hybrid vision pipelines (OpenCV, feature extraction, template matching) where deep learning isn't the right tool
• Work with imaging hardware — cameras, lenses, lighting, and optics — to optimize image quality at the source
• Curate and annotate datasets; design augmentation and synthetic-data strategies for rare defect types
• Optimize models for real-time inference (TensorRT, ONNX, quantization) on edge/GPU hardware
• Set and tune decision thresholds balancing escape rate vs. overkill; monitor models in production and retrain as processes drift
• Partner with manufacturing, process, and equipment engineers to integrate vision systems into production lines and MES
• Build evaluation harnesses and drive continuous improvement of inspection accuracy and throughput **What You'll Bring**
• 3+ years of experience in computer vision, with hands-on work in deep-learning based AI systems
• Strong Python skills; production experience with PyTorch or TensorFlow
• Solid grounding in both classical image processing (OpenCV) and modern deep learning (object detection, segmentation, anomaly detection)
• Experience taking models from prototype to production: real-time inference, latency/throughput constraints, edge deployment
• Experience with dataset creation, annotation workflows, and imbalanced / rare-class problems
• Comfortable working in a manufacturing environment; you care about false positives as much as accuracy
• Strong English reading and writing skills
• Nice to Have**
• Familiarity with imaging hardware fundamentals — cameras, lenses, illumination — and how they affect model performance
• Experience with semiconductor, PCB/electronics, display, or precision-component inspection
• MLOps experience: model versioning, CI/CD for ML, production monitoring