Skip to content

Deep Learning Brand Recognition, using client-server architecture for low resource consumption & easy deployment.

License

Notifications You must be signed in to change notification settings

codeadeel/Brand-Recognition

Repository files navigation

Brand Recognition

DockerHub CI

Brand Recognition Macro Architecture

The subject repository is responsible for Brand Recognition. Final classification is implemented as server-client architecture. Scenes under consideration are segregated using features extraction and matching using deep learning tools. This system can be applied on multiple streams for identification of different scenes, already available in feature store. Simply explained, if a scene to be tracked is available, it can be identified on live stream.

The model used in this architecture is ResNet-18, on Pytorch framework.

For extensive documentation, please check wiki.

About

Deep Learning Brand Recognition, using client-server architecture for low resource consumption & easy deployment.

Topics

Resources

License

Stars

Watchers

Forks