Job Description
We are looking for a Senior Computer Vision / Image Processing Engineer to design and develop end-to-end image processing and computer vision pipelines for advanced, real-world applications. The role emphasizes segmentation, detection, image quality optimization, and close collaboration with ML and product teams in an R&D-focused environment.
Roles and Responsibilities
- Design and implement end-to-end image processing pipelines, including preprocessing, segmentation, and object detection.
- Develop software-based color correction and image enhancement techniques using statistical and machine learning approaches (e.g., Gray World, SVM, SVR).
- Apply morphological operations (erosion, dilation, interpolation) for noise reduction and artifact removal.
- Implement and optimize image segmentation techniques, including:
- Classical methods (Active Contour / Snake models, Watershed)
- Deep learning–based architectures (UNet, UNet++, or similar)
- Customize, fine-tune, and optimize real-time object detection models, preferably from the YOLO family, for domain-specific use cases.
- Work with large-scale image datasets, contributing to annotation strategies, quality guidelines, and tooling.
- Collaborate closely with ML engineers, researchers, and product teams during experimentation, validation, and deployment phases.
- Document experiments, models, and pipelines to support reproducibility and knowledge sharing.
Required Skills
- Strong hands-on experience in Computer Vision and Image Processing.
- Proficiency in Python, OpenCV, and modern deep learning frameworks (PyTorch / TensorFlow).
- Proven experience in image segmentation and object detection.
- Strong understanding of image quality correction, enhancement, and optimization techniques.
- Experience working in R&D or experimentation-heavy environments with rapid iteration cycles.
- Strong problem-solving and analytical skills.
Nice to Have
- Experience in medical imaging, healthcare AI, or clinical data processing.
- Prior exposure to early-stage R&D or prototyping environments.