Computer Vision Research Engineer - Deep Learning
Parkopedia, United Kingdom
Job Description
Parkopedia was founded with the mission of being able to answer any parking question, anywhere in the world. Today, Parkopedia is the world’s leading parking services provider used by millions of drivers and organisations ranging from Apple to Volvo.
The goal of the Computer Vision team is to develop novel methods for providing static and dynamic parking related information worldwide. We use Computer Vision and Machine Learning/Deep Learning to find parking and predict availability for parking locations in thousands of cities. We are looking for a computer vision and deep learning expert to apply state of the art methods on images/video/point cloud data.
As a Computer Vision/Deep Learning expert you will design, implement and test algorithms for object detection, recognition and tracking, semantic segmentation, stereo and monocular depth estimation, etc. It is a unique opportunity to get hands-on experience with state-of-the-art technology on video sequences and imagery collected globally, and you will work alongside other computer vision and machine learning experts in the growing organisation.
We are part of a global team but are based in the London Office, which is a short walk from London Bridge.
Responsibilities
- Experiment with, research and develop novel computer vision algorithms to extract parking information
- Design, implement and extend state-of-the-art deep learning models
- Document and clearly articulate your work, to both clients and colleagues
- Evangelise and share knowledge through writing blog posts or speaking at meetups
- Contribute to our deep learning workflow and infrastructure
- Working alongside an annotation team by writing highly detailed instructions for manual annotation tasks and subsequent work validation
- Be a great member of our team, while being able to work autonomously
Requirements
You will have:
- At least 3 years of experience working on solving Computer Vision problems with deep learning libraries such as TensorFlow (ideally) or PyTorch
- Familiarity with a wide range of DL approaches and experience with at least one of the following topics: Object detection, Object tracking, Segmentation from sources such as images, videos or point clouds
- Strong programming skills in Python and comprehensive knowledge of its ML/CV ecosystem (e.g. Pandas/Geopandas, Numpy, OpenCV, Scikit-Image, Scikit-Learn)
- Passionate about writing well-structured and well-tested Python code
- A strong mathematical background covering linear algebra, calculus and optimisation techniques, from a MSc or PhD in a highly mathematical field
- A desire to keep up with the latest developments in the field through reading and talking about the latest publications within a knowledge sharing ecosystem
- Good working knowledge of SQL
- Proficiency working with *nix-based environments
Preferably you will also have:
- Industry experience in scaling Deep Learning projects from prototype to production
- Experience in setting up and maintaining ETL pipelines
- Understanding of traditional CV methods and multi-view geometry
- Evidence of personal interest in the subject, through for example: published papers, open source repositories or Kaggle achievements
Benefits
Parkopedia is committed to building a great work environment for all our employees. Here are just a few of the benefits that we offer:
- Unlimited annual leave - yup, time off is as important as time in the office, we all need to unwind and recharge our batteries!
- Flexible working hours
- Training allowance
- Fresh fruit, tea, coffee
- Annual company retreat
- Private Medical Insurance
- Regular team lunches and socials
- Time off for volunteering
- Cycle to work scheme
- Gym membership
- Eye care and flu vouchers