Machine Learning Engineer

About the Job

We’re looking for Machine Learning Engineer for computer vision to join our high-impact applied research team.

This role will play a key part in pushing the boundaries of our real-world video analytics platform, focused on improving traffic safety and transport insights through state-of-the-art computer vision and deep learning techniques.

You’ll lead hands-on development of detection, segmentation, and multi-object tracking models designing and testing innovations that will directly improve outcomes for traffic engineering and transport policy. This is a deeply technical role at the intersection of AI research and real-world application.

You’ll be working alongside a close-knit team of engineers and researchers, with significant autonomy, access to real-world datasets, and opportunities to publish and share your work. The position is ideal for someone with strong research experience and a passion for applied impact.

You Should Have

• A PhD in Computer Science, Applied Mathematics, Robotics, or a related discipline
• (Alternatively, a Master's/Bachelor's with a strong applied AI research track record)
• Proven expertise in object detection, segmentation, and multi-object tracking
• Experience with complex video analytics scenarios (e.g. occlusion, re-identification)
• Proficiency with PyTorch or TensorFlow, and core libraries like OpenCV and NumPy
• Demonstrated experience deploying and optimising deep learning models in real-world settings
• Experience with video artefact denoising and performance improvement techniques
• Strong understanding of camera calibration and spatial measurement workflows
• Ability to independently design and run ablation studies and field validations
• Excellent communication skills and ability to collaborate across technical domains

Nice to Have

• Experience with ONNX, TensorRT, or Triton for GPU-based inference
• Familiarity with transport or traffic datasets and traffic engineering metrics
• Exposure to MLOps tools like MLflow or Weights & Biases
• Research background in Transformer-based architectures for video analytics
• Publication record or open-source contributions in computer vision or applied AI

Key Responsibilities

• Investigate and benchmark model architectures for detection, segmentation, and tracking
• Enhance and optimise multi-object tracking pipelines using DeepSORT, ByteTrack, or Transformer-based approaches
• Evaluate and implement new model strategies to improve real-world video performance
• Optimise inference runtime and end-to-end latency across the video processing pipeline
• Apply advanced denoising techniques to mitigate artefacts from low-resolution or compressed footage
• Lead field validation of models using drone, fixed, and varied camera types
• Contribute to post-processing, trajectory smoothing, and spatiotemporal error reduction
• Support spatial calibration workflows for accurate measurement of real-world traffic data
• Collaborate with cloud and software engineering teams for scalable model deployment
• Translate AI outputs into actionable transport safety insights
• Document experimental findings and contribute to internal and external publications

The Benefits

• Join a mission-driven team using AI to make transport safer and smarter
• Work on impactful projects with real-world deployment and policy relevance
• Hybrid or fully remote work options
• Access to cutting-edge tools, models, and real-world datasets
• Collaborative, multidisciplinary environment that values innovation and rigour
• Opportunity to publish, present, and contribute to open-source research

How to Apply

If this opportunity sounds like a strong fit, we’d love to hear from you. Please apply, and feel free to reach out via email at [email protected] or call Kent Sin on (08) 6146 4464.