WEEK 1- Brand Recognition Using Machine Learning from Google Street View Images

Theme: Predicting the Brands

Team Members : Fırat Kaya, Hülya Şermin Karakaş

Here’s a question that perhaps needs to be asked, but hasn’t been asked enough. Did you ever think, how many different brand you are using in your daily life? Or, did you looked at the environment around you?

Every place of our lives is full of brands nowadays. In addition, it will rapidly increase day by day. Because of that, we plan to focus working on brands.

About Our Project

In our project, we are planning to recognize brands by reading letters. We will train the program by alphabet dataset which has lots of different fonts, then operate it with google street view images. While doing that, we are planning to detect borders of the signboards and tag them by their brands. If we could finish the recognition by letters earlier than we expect, we could improve our algorithm by recognition of the brand logos.

The first and most important step to start a machine learning project is collecting the data. Thus, we will collect data from some character dataset and Google street view images. If we will need, we also plan to use flicker dataset.

Related work