Address
Form of workThe Algorithm Development Team brings cutting edge research in areas such as fatigue and distraction detection into occupant monitoring solutions. Take an example of fatigue detection, which is conceptually simple – just detect when someone looks tired, right? To develop this feature, there are many underlying details that need to be understood. For example, how do you collect a dataset safely? What is the true drowsiness level? How to account for individual behavioural differences? Algorithm Development Team members work closely with Human Factors Research Scientists, Data Acquisition Specialists and Machine Intelligence Researchers in order to make progress on these kinds of questions in order to develop features.
Aside from developing new features, the Algorithm Development Team also helps improve existing features that are currently used by customers. For example, making improvements to handle specific real-world corner cases, improving performance, etc.
About the opportunity
The Machine Learning Scientist will work directly with internal stakeholders to develop occupant monitoring features using state-of-the-art machine learning methods (e.g. NN, CNNs, RNNs), traditional computer vision techniques, or often a combination of both.
One aspect of feature development includes collaborating with Human Factors Research Scientists and Data Acquisition Specialists to help collect relevant, truthed datasets. Also required is working closely with our Performance Analysis Team to ensure that we can measure and validate the performance of the algorithms.
You will be responsible for:
- Developing proof of concept algorithms and prototypes for features that will be continually evolving in Python and C++.
- Writing clear documentation to help define features, including use cases, dataset specifications, performance targets, etc.
- Making use of state-of-the-art machine learning techniques (eg NN, CNN, RNN) utilizing large datasets to train algorithms.
- Providing guidance to embedded software engineers to enable algorithms to be used on specific embedded hardware platforms.
- Assisting with business pursuit activities such as working with potential partners, assisting customers to evaluate our technology and fielding questions from customers.
- Responding to customer-reported issues when required in order to fix problems or limitations encountered in real-world conditions.
We value the breadth and depth of experience that our people bring to their work. We are seeking people with a good mix of technical competency and personal capability. Specifically, we are looking for people with:
- A Masters or PhD (awarded or currently completing) in Computer Vision, Machine Learning, or equivalent industry experience
- Strong theoretical understanding of machine learning concepts such as CNNs, RNNs, etc.
- Experience with machine learning frameworks such as TensorFlow, PyTorch or similar.
- Experience writing code in Python and/or C++, ideally in a commercial environment.
- Evidence of producing high-quality research outcomes and in disseminating effectively to academic and industry settings.
- Understanding of traditional computer vision and/or image processing techniques.
- Experience working on research in a commercial environment.
- Experience with practical data science or statistical analysis.
- Experience explaining and discussing technical topics with varied audiences. For example, talking with customers, internal engineers, or highly technical peers.
To apply, please upload your current resume and cover letter outlining your interest in the role. If you are successful for this opportunity a police check is required.
About Seeing Machines
Seeing Machines is a world leader in driver-machine interaction. We harness human factors science to create artificial intelligence (AI) technology that observes the driver's attention – reliably, unobtrusively, and in real time – and intervenes seamlessly when necessary.
Specialised computer vision algorithms underpin Seeing Machines' core camera-based driver monitoring technology. Algorithms allow us to precisely track eye gaze, head position and pupil size while our state-of-the-art AI technology analyses the data to quickly and accurately detect driver drowsiness, distraction and microsleep levels. We work with some of the world's leading brands to deliver this technology, helping keep drivers and operators engaged across commercial transport and logistics, automotive, aviation, rail and mining industries.
In automotive, we enable safer Automated Driving (AD) solutions and Advanced Driver Assistance Systems (ADAS) – including the world first camera-based hands-free driving feature, GM Super Cruise. In aviation, our advanced gaze tracking technology understands how pilots interact and monitor instruments – leading to better training and safer operations.
For commercial transport and fleet operators, the Seeing Machines retrofit driver monitoring technology provides real-time intervention for drivers to mitigate the risks associated with driver fatigue and distraction.
For more information, visit: https://www.seeingmachines.com/technology/
