The candidate will join a small, agile group of computer vision engineers and provide expertise related to real-time object detection and segmentation to support a variety of cutting-edge Augmented Reality applications.
The candidate will be doing research, prototype and survey different DL networks/algorithms that are the current state-of-art and model optimization techniques (e.g., compression, quantization).
The candidate will also be required to develop and contribute to the implementation of various computer vision related applications as needed.
MUST HAVE SKILLS:
• 5+ years experience with development and implementation of computer vision or deep learning algorithms.
• Significant, demonstrable experience with state-of-the-art deep learning networks in
o Real-time Object Detection techniques for 2D/3D objects,
o Semantic/Instance segmentation,
o Hand gesture recognition,
o Human pose estimation,
o Training dataset design, augmentation, and validation,
o Classification/Segmentation performance metrics.
• Experience in working with large data sets and developing ML infrastructure pipelines.
• Strong experience in Neural network optimization (pruning, quantization and compression).
• Strong experience with python, C++, and OpenCV required.
• Strong expertise with DL frameworks like TensorFlow, Pytorch, Mxnet, Onnx etc.
• Strong expertise with software development lifecycle - coding, debugging, optimization, testing, integration
• Experience with DL inference accelerator toolkits like TVM, Openvino, TensorRT, PlaidML (optional)
B.S/M.S in Computer Science or equivalent college degree.