Deep Learning optimizations developer

Job Descriptions    

 Working as a part of a cutting-edge algorithms team that is researching state of the art computer vision challenges with combination of classical and DNN approaches, in the autonomous car world. Including camera calibration, stereo vision, sensors fusion, detection, and segmentation tasks.  

In this role you will be responsible for accelerating and improving neural networks for real time applications. Working with large data and video streams. Leading the implementation of the feature by SW team. Occasionally writing patents and scientific articles and presenting at conferences.  

• Move models from research to deployment 
• Optimize runtime efficiency of DL models in production 
• Porting DL models to edge devices such as NVIDIA Jetson 
• Speed up and optimize training  
• Build data pipelines   
• Support multiple servers use for training  
• Automation tools for simulation and evaluation of performance  
  
  

• 3 years at least of experience of programming (must)  
• Excellent programming skills in Python or C/C++  (must) 
• 3 years at least experience of working with Pytorch (must) 
• Experience in TensorRT / CUDA / network acceleration methods (must) 
• Experience working with dockers / Conda environments (must)  
• Experience in neural networks optimization and porting to edge device (advantage) 
• Experience with Computer vision: stereo/ registration/ denoising/ augmentations, detection, segmentation, etc. (advantage)   
  
  

• M.Sc. or PhD in Computer Science / Electronic Engineering / Math – must  
• B.Sc will be considered in rare cases 
  
  

•          Willing to commute to Park Ha'Mada Rehovot (accessible by train). Hybrid office/home 3/2 policy.