[Last week — Predicting the pieces of furniture in living rooms]

Group Members: Mohammed ALI, Aybüke Yalçıner, HaticeAcar.

We were so keen to find a solution with high accuracy. It’s our aim in this project. we do not give up trying to optimize our solution since we see the first result. We have tried all that we can do in order to come up with a with good result. By changing the attributes and adding more hidden layers. We increased the accuracy from 69.15.15 to 75.28. As you see in the table below.

Save our trained model

After trained our model multiple times, we saved it as a trained model to use it whenever we want. Training time it takes nearly 3 hours with 100 epochs. Trained model can be used to directly to predict the furniture.

Strengths and weaknesses of our project

We are not just doing the project but we are evaluating our work at the same time. Like most machine learning projects, our project has achieved most of what we planned to do. We are satisfied with our all works. Limitation of your resources and collecting data for each object manually cost us much time. It takes much time to run because of dealing with images.

Future work

We are planning to add more objects, for example, we are going to add all object in sleeping rooms, kitchens and other stuff that are part of each home. By adding more objects and images for each one. We are also going to use the power of cloud computing instead our laptops

Our dreams

We believe that applying technology in our life can help us to avoid doing frequent daily work to save our time and let us do more challenging works or spend time with our family. And we are close to achieve that more than any time before.