Senior Autonomy Engineer - Deep Learning

Skydio · Defense · Zurich, Switzerland · R&D

Senior Autonomy Engineer focused on designing, implementing, and optimizing deep learning solutions for real-time object detection, tracking, segmentation, and optical flow estimation on drones. The role involves leveraging state-of-the-art methods like unsupervised learning and foundational models, curating synthetic data, and refining models for low-latency embedded hardware, with a strong emphasis on computer vision and robotics applications.

What you'd actually do

  1. Design and implement deep learning solutions that solve detection, tracking, segmentation, and optical flow estimation tasks in real-time on Skydio drones
  2. Leverage state-of-the-art methods in unsupervised learning, few shot learning, and foundational models for data efficient ML
  3. Curate and enhance synthetic data that powers our deep learning algorithms along with massive amounts of structured video data
  4. Refine and optimize models for low-latency on embedded hardware
  5. Characterize and quantify the performance of the vision systems

Skills

Required

  • Deep learning
  • Object detection
  • Object tracking
  • Segmentation
  • Optical flow estimation
  • Unsupervised learning
  • Few shot learning
  • Foundational models
  • Data efficient ML
  • Synthetic data generation
  • Low-latency model optimization
  • Embedded hardware deployment
  • Computer vision
  • Python
  • C++
  • Software engineering
  • Scientific paper comprehension

Nice to have

  • Both Python and C++
  • Robotics

What the JD emphasized

  • real-time deep networks
  • object detection and tracking
  • motion prediction
  • flow estimation
  • total scene understanding
  • massive amounts of structured video data
  • unsupervised learning
  • few shot learning
  • foundational models
  • data efficient ML
  • synthetic data
  • low-latency on embedded hardware
  • vision systems
  • creating and deploying deep learning models
  • curating synthetic and real-world image datasets
  • prototyping, training, optimizing, and deploying deep neural networks

Other signals

  • real-time deep networks
  • object detection and tracking
  • motion prediction
  • flow estimation
  • total scene understanding
  • massive amounts of structured video data
  • unsupervised learning
  • few shot learning
  • foundational models
  • data efficient ML
  • synthetic data
  • low-latency on embedded hardware
  • vision systems