Week 3- DASM - Dataset Building and Model Selection

Hi everybody, I will tell you the latest news about our Damage Assesment System For Imagery Data (DASM) project. This week’s topic is final situation dataset building and the possible models we can use in our project.

In our previous blog, we said that we downloaded the data and removed the repeating images from the data set. In this week, we labelled the downloaded images according to their damage assessment level, We are using the DataTurks image labeling tool, which is a web-based tool that facilitates the collaborative labeling process.

We have not yet decided which methods to use in our project, but the methods we plan to use are like these:

1. The bag of visual words (BOVW) model is one of the most important concepts in all of computer vision. We can use the bag of visual words model to classify the contents of an image.

An example of constructing a bag of visual words from an image

2. Convolutional neural network(CNN) is a neural network that has one or more convolutional layers. We can use the convolutional neural network to classification of the images.

Thank you so much for reading, and stay tuned for the coming blogs!.