Website MoTek Technologies
We’re looking for a Research Engineer with a deep passion for deep learning and computer vision. You will research, design, implement, optimize and deploy models and algorithms that advance the state of the art in contextual understanding in a smart home environment. You will need to be able to demonstrate a deep knowledge and expertise in computer vision and deep learning as well as the ability to produce production quality code and models while working in a fast-paced dynamic team.
The Role:
Design, implement, train, evaluate and optimize novel perception algorithms based on state-of-the-art deep learning techniques, such as convolutional neural nets (CNNs) and deep neural nets (DNNs).
Solve problems related to object detection, scene classification/understanding, segmentation, pose estimation and recognition with the goal to be deployed in cloud-connected and embedded consumer device products.
Define and perform experiments leveraging relevant large scale datasets to identify and tune viable proof-of-concept approaches. Transfer and optimize proven approaches to run on cloud, embedded processors, or hybrid targets.
Qualifications:
A solid understanding of and experience with Deep Learning, in particular with design, training, evaluation and optimization of convolutional neural net (CNN) architectures in the context of object recognition, scene classification and scene segmentation.
Experience with at least one major deep learning framework (such as Caffe, Theano, Torch, Tensorflow). Experience with Tensorflow or Caffe preferred.
Self-motivated learner who keeps up-to-date with the current state of computer vision and deep learning techniques/architectures.
A solid background in machine learning/classification concepts.
A solid background in computer vision/image processing concepts.
Minimum of 3 years experience in algorithm implementation in C++ and Python.
Hands on experience with computer vision operating on real-world datasets.
Strong analytical skills and mathematical foundation.
Good understanding of computer systems / architecture trade-offs.
Comfortable in a Linux-based environment.
Excellent verbal and written communications skills.
Ability to work independently, without direct supervision.
Strong problem solving skills and ability to learn quickly.
Minimum Degree: BS/MS/PhD in Computer Vision, Computer Science, Data Science, Machine Learning, Robotics, or related field
Desired Qualifications:
Experience with CUDA GPU programming.
Experience with Deep Net compression techniques (both size and speed).
Experience with embedded processor platforms.
Experience with mobile GPUs / OpenCL / OpenGL / shader programming.
Experience with robotics or other embedded real-time systems.
Experience with RGB-D sensors and algorithms.
Background in SLAM, 3D reconstruction, motion planning.
Experience with IoT protocols and products.