People love to feel popular on social media, with more likes or upvotes on their pictures making them feel happier. We want to help boost people's self-esteem by helping them judge their pictures before uploading. This prevents people from uploading possibly bad-quality pictures.
What it does
Analyses the picture and categorises it into either good, bad, or neutral. Our app also provides feedback on why that picture is either good or bad.
How we built it
We used Face API to read the image and get back data such as how blurry the picture is. Using these attributes, we set thresholds to determine if the photo was good. We also tried to incorporate custom vision from Microsoft Cognitive Services, training the system on good and bad profile pictures. This would aid the algorithm in refining what defines a good picture.
Challenges we ran into
Finding suitable datasets for training the model.
What we learned
How to make use of Microsofts Cognitive Services.
What's next for ProProPicPickr
Advancing the classifier and algorithms. Adding an option to directly set as your profile photo.