Computer Vision Engineer

As a computer vision engineer, you will play a pivotal role in researching, developing, and implementing spatiotemporal algorithms for detecting and delineating lesions in real-time during upper gastrointestinal endoscopic procedures. You will collaborate with a multidisciplinary team of AI researchers, gastroenterologists, and software engineers to enhance both spatial and temporal accuracy in lesion detection. This role involves designing preprocessing pipelines, optimising models, and ensuring low-latency, high-performance deployment in clinical settings.

Responsibilities:

• Exact ownership and depth will depend on whether you are joining as an engineer or an intern.
• Interns will primarily work under close mentorship, while engineers will own features end-to-end from research to deployment.

Algorithm and Model Development:

• Research, prototype, and implement spatial and spatiotemporal models (e. g., CNNs, 3D CNNs, LSTMs, and Transformer-based models) for lesion detection and tracking.
• Design and optimise preprocessing pipelines (frame extraction, enhancement, noise reduction) for real-time video feeds.

Integration and Deployment:

• Integrate trained models into our inference pipeline for seamless, low-latency deployment on clinical hardware (Jetson-based devices).
• Collaborate with software engineers to bridge C++ modules (GStreamer/other low-latency frameworks) and Python inference wrappers.
• Containerise and deploy vision pipelines with Docker on Linux medical workstations; manage CI/CD for model updates.

Performance Benchmarking and Optimisation:

• Benchmark model performance (accuracy, speed, robustness) on endoscopic datasets.
• Profile and optimise code for GPU/CPU inference; implement quantisation or pruning as needed for resource-constrained environments.

Regulatory Compliance and Collaboration:

• Work alongside regulatory and QA teams to document algorithms and validation processes per FDA/ISO 13485 standards.
• Participate in design reviews and risk assessments and prepare technical documentation for audits.

The core requirements for the job include the following:

For Computer Vision Engineer (Full-Time):

• Bachelor's, Master's, or PhD in Computer Science, Electrical Engineering, or a related field with a focus on AI/computer vision.
• Solid understanding of computer vision fundamentals: image processing, feature extraction, neural architectures.
• Hands-on experience with deep-learning frameworks (TensorFlow, PyTorch, or Keras).
• Production experience with deep learning inference pipelines.
• Experience designing and optimising real-time inference pipelines.

For Computer Vision Intern:

• Currently pursuing or recently completed a degree in CS/EE/AI.
• Strong fundamentals in computer vision and deep learning.
• Proficiency in Python; C++ exposure is a plus.
• Strong curiosity and ability to learn fast.

Nice to Have (Both Roles):

• Experience working on medical imaging or regulated software (FDA, CDSCO, ISO 13485).
• Familiarity with GStreamer or low-latency video frameworks.
• Experience with Docker, Linux, and Git-based workflows.
• Exposure to edge deployment, hardware acceleration, or model optimisation.