AI Researcher

Birds.ai is looking for a full-time AI researcher to join our team!

Inspectors are using drones to look for defects on large structures such as solar panels and wind turbines. These drones collect thousands of images. Going through these images manually is time-consuming, boring and very expensive.

At Birds.ai we use Artificial Intelligence to create Artificial Inspectors that check all of these images to see if they contain a defect instantly and automatically.

The Artificial Inspector selects images that show damage or defects, it can precisely pinpoint the location of the defect and it can even automatically deliver inspection reports.


Mission

  • Become the best at automatically finding objects in images.

  • Use Artificial Intelligence (deep learning) to find defects on assets such as wind turbines in images collected by drones.

  • Build AI that can easily be scaled to different business cases (for example: inspection, agriculture and wildlife conservation).

  • Turn Artificial Intelligence into Artificial Inspectors that will give advice on repairs and predict failures from happening at all.


Tasks

  • Find defects: research and implement ConvNets for defect detection.

  • Find objects: research and implement ConvNets to detect any object in aerial imagery (for example: defects, bushes and rhinos).

  • Train models with few examples: research ways with which we can quickly fine-tune our existing models to perform well on new types of data with little (< 100 examples) training data.

  • Keep on learning: research ways in which we can fine-tune models without forgetting previously learned classifying parameters to speed up training and enable the detector to learn while detecting.


You...

  • have experience with deep learning for practical applications (show or tell us what you did!).

  • are able to quickly create prototypes to evaluate your ideas and iteratively improve them into working applications.

  • can solve challenges independently.

  • are an excellent project planner: you can take a long-term goal and break it up into achievable steps.

  • worked with a machine learning framework (e.g. Caffe, Torch, TensorFlow, etc.).

  • are a resident of the EU, or eligible for a Dutch Working permit and are willing to move to the Netherlands.