Vidrovr Inc. seeks Research Software Engineer in New York, New York. S/he will take part in building a unified framework for processing and combining multimodal sources of information for video understanding. This framework will be built using python, numpy, scipy, tensorflow, opencv and other scientific python libraries. Some portions of the pipeline may be written in C++ (if deemed necessary). In addition, s/he will develop various forms of deep learning models implemented in Tensorflow for video classification, leveraging recurrent neural networks (RNNs) such as LSTM/GRU, memory nets like differentiable neural turing machines, hierarchical neural nets and others. The Research Software Engineer will work on training high speed video object detection models like SSD, Faster RCNNs. S/he will develop various aggregation and data modeling pipelines built atop modern graph databases (Neo4j) and key val stores (RocksDB). Furthermore, the Research Software Engineer will work on more classical approaches building shot/scene detectors, OCR pipelines leveraging some modern detector implementations.
Other specific duties include the following:
The Research Software Engineer must have, at a minimum, a Master’s degree in Engineering and two years’ experience working in a systems or networks focused technical engineering role. Email resume to firstname.lastname@example.org.