Standing in front of the trash bin for few minutes standing with empty cups and used packets, trying to figure out which bin to use out of the three.
What it does
Trash-Ed scans the trash image and predicts which bin the trash should be disposed off, to help the environment.
How we built it
We created an Android application that used google vision api. We trained our model that segregates trash into different categories using dataset of approximate 4k images.
Challenges we ran into
Google authorization for autoML has been a challenge.
Accomplishments that we're proud of
We were successful in training our dataset using 4k images with adequate accuracy
What we learned
We learnt the how the google vision autoML works.
What's next for Trash-Ed
-Geo-tagged bins to get the nearest location of the trash bin -Object detection of each trash object with details. -Add a device on dustbin that senses the image and predicts which bin to use.