IDEXX is looking to develop image analysis algorithms for veterinary clinical instrument platforms, specifically microscopy algorithms for their inVue platform, to transform cytology analysis in veterinary clinics.
Requirements
- Foundational knowledge of both classical computer vision techniques and modern deep learning approaches.
- Proficiency in Python and experience with deep learning frameworks, particularly PyTorch or TensorFlow.
- Experience designing, training, and evaluating CNNs or other neural architectures for image classification, segmentation, or detection (through coursework, projects, internships, or research).
- Exposure to model optimization and deployment considerations for resource-constrained environments.
- Understanding of challenging discrimination problems in imaging (e.g., subtle feature detection, class imbalance).
- Solid foundation in experimental design including dataset preparation, validation strategies, and performance evaluation.
- Familiarity with cloud platforms, particularly AWS, is a plus.
Responsibilities
- Develop and implement computer vision algorithms combining supervised deep learning (CNNs, transformers), unsupervised machine learning, and classical techniques (feature extraction, morphological operations, edge detection) for discrimination tasks.
- Assist in the research, prototyping, and implementation of computer vision solutions for medical imaging and cellular analysis applications.
- Support the development and validation of image processing pipelines that perform reliably across diverse imaging conditions and sample variations.
- Work with the team to implement and optimize computer vision models for deployment on embedded devices and NVIDIA GPUs.
- Conduct experiments and performance evaluations to validate algorithm performance, including benchmarking and statistical analysis.
- Collaborate with cross-functional teams including engineers, clinical experts, and product stakeholders to understand and address clinical requirements.
- Contribute to technical documentation for computer vision development, including algorithm design and validation methodologies.
Other
- This role will be hybrid in our Westbrook, ME office, onsite 8 days per month.
- Unfortunately, we are unable to provide relocation assistance or sponsorship for this role.
- M.S. in Computer Science, Electrical Engineering, Data Science, or related field, or B.S. with relevant internship/project experience in computer vision.
- Strong analytical and problem-solving abilities with attention to detail.
- Interest in or exposure to microscopy imaging, digital cytology, or histopathology is a plus.
