Currently there are several libraries, models and datasets available for Computer Vision research. However its still difficult to apply a model to your own dataset without writing a non trivial amount of glue code. There is a need for a drop-in solution that comes with UI, algorithms and pre-indexed dataset ready to use out of the box. Deep Video Analytics aims to be the platform to enable this. Using Deep Video Analytics you can quickly index images, videos for visual search, detect objects, and soon recognize faces and incorporate large pre-indexed datasets via fast approximate nearest neighbors.
1 comment
[ 0.22 ms ] story [ 8.0 ms ] thread