Principal Computer Vision Engineer
| Company | CHEP |
|---|---|
| Location | Santa Clara, CA, USA, Mississauga, ON, Canada, Madrid, Spain, Manchester, UK |
| Salary | $Not Provided – $Not Provided |
| Type | Full-Time |
| Degrees | PhD |
| Experience Level | Senior, Expert or higher |
Requirements
- PhD in Computer Vision, Computer Science, Engineering, or related field
- Expert in using AI, deep learning, and traditional computer vision for facial recognition, person re-id (reidentification) or unique identification of objects
- Expert in the use of model optimization techniques such as quantization, pruning, and knowledge distillation
- Expert level proficiency with computer vision and deep learning toolkits including OpenCV, Pytorch, ONNX, TensorRT
- Expert level proficiency with Python
- Proficient with creation, indexing, and searching embeddings, latent space, and vector representations of data
- Proficient with MLOps tools such as bitbucket/git, DVC, MLflow, JIRA, jupyter, docker
Responsibilities
- Lead research of nascent computer vision projects including literature review, method evaluation, data collection & exploration, supervision of annotation efforts
- Lead review of established computer vision projects and integrate state-of-the-art research and software to improve performance and deliver new features
- Leverage large image sets to construct and evaluate ML/AI prototypes while minimizing resource requirements
- Optimize computer vision models for deployment on resource-constrained edge systems
- Monitor and fine-tune production ML systems using techniques such as drift monitoring and active learning
- Guide computer vision team discussions by providing insight on potential approaches, assist with troubleshooting and problem solving
- Assess project objectives and tactical plans to ensure successful delivery and provide recommendations to team and project leadership
- Keep current on the latest findings from researchers in relevant computer vision fields
Preferred Qualifications
- Proficient with multi-node, multi-GPU training of deep neural networks using model and data parallelism
- Proficiency with Go, C/C++
- Proficiency with Tensorflow