I am very much influenced by AI when i saw Facebook image processing then i think that if it is used to detect background object of the user upload image and video it will provide benefits to Facebook as well as User.

What it does

This program takes Image and Video File as an input and provides output with their names. It can able to Detect 80 objects for now with high accuracy.

How I built it

I built it using python and Pytorch mainly however other different modules are used for better Result .

Challenges I ran into

Since i don't have NVIDIA graphic it is very difficult to train and to stream video file. Instead of using CUDA i used CPU to compile my program which is slower to get result.

Accomplishments that I'm proud of

After the accomplishment of this program i am very proud to see the accurate result ,even though it can detect the object from video file accurately.

What I learned

After the completion of this program i learned how Pytorch works in machine learning and it's importance for Future AI world.

What's next for Object Detection and Tracker

This program is just to detect the background object however if Facebook application wants to used it Recommendation feature should used after the detection of objects. If we are able to recommend the right product it would increase the business and have Positive impact on the society.

Share this project: