[Week-1 Object Detection and Room Classification with Deep Learning]
BBM 406 Fall 2018 Semester Project
Team Members:
- Kaan Mersin
- Ayca Meric Celik
- Ahmet Tarik Kaya
There are a lot of different types of objects and tools used in regular houses and most of them can be used to classify the room. In this project, our aim is to detect important objects like furniture, curtains, lightings, carpets, etc. We will also label the rooms by their types with this information.
We will give details about our project below.
Part 1
In the first part of our project, we will detect multiple objects in room pictures. We will try different methods and approaches for this purpose and find the most efficient and accurate one. We have already decided on some of these.
Part 2
We will try to classify the room by using the detected objects such as living room, bedroom, bathroom, kitchen, hall, etc.
Our Datasets
The first dataset which we mentioned in our project proposal was an RGB-D dataset for object classification. However, we decided that working with an RGB-D dataset will be a difficult task at the recommendation of our instructor.
So we will search for an RGB dataset for our project this week. If we cannot find a suitable one, we will try to work with RGB-D datasets. After that, we will implement our first model for Part 1.
We already have a dataset which contains labeled room pictures. We will use it for the second part of our project in the future.
Resources
