About the job
Research and prototype anomaly/defect detection methods for high-resolution industrial vision data. Design and implement proof-of-concept ML solutions from research ideas, train, evaluate, and analyze computer vision models under real-world data constraints. Iterate on models using quantitative metrics and qualitative failure analysis.
Responsibilities
Research and prototype anomaly/defect detection methods for high-resolution industrial vision data
Design and implement proof-of-concept ML solutions from research ideas,
Train, evaluate, and analyze computer vision models under real-world data constraints
Iterate on models using quantitative metrics and qualitative failure analysis
Qualifications
Minimum
Working towards a MS or PhD in AI/ML, CS, EE, or related field
Preferred
Currently pursuing or recently completed a Master’s or PhD in AI/ML, CS, EE, or related field
Hands-on experience implementing and modifying modern vision models (CNNs, Transformers) rather than only using pre-trained models.
Demonstrated experience training at least one vision model end-to-end (data preparation → training → evaluation), documented via coursework, research, or projects.
Strong Python skills; experience with PyTorch or TensorFlow
Experience with unsupervised or weakly supervised learning is a plus
Exposure to multimodal or vision-language models is a plus
Strong analytical and problem-solving skills