Intelligent Logo detection on video frames

What it does

Train a network using the data set provided by client. Learn if logo is present in the image or not. Then predict the presence of logo in the training set.

How we built it

We used Convolutional Neural nets, built using Tensorflow. the UI is built using Django

Challenges we ran into

  1. There are very less training examples for some logos.
  2. Computational resources to train Conv Nets.a

Accomplishments that we're proud of

The network is general enough to provide an output whether a logo is present in some region of image or not. It does not require any predefined logo to chose from.

What we learned

Conv Nets, Team spirit

What's next for blackhats

Improve the model

Built With

Share this project: