Integration trash-ai in citizen science project
Opened this issue · 5 comments
Hi Everyone,
In terms of my master thesis I am developping a mobile application to gather data about the waste consumption on the island of Aruba. I am trying to integrate the trash-ai algorithm, so that users just have to take a photo of their trash and the trash can be recognized by the algorithm (and than later saved in a database).
I am a newbie in software, and certainly in docker, AWS, ... . I just want to communicate with the trashai API. So that I can send post requests (with the photo) and get a response that has the information of the recognized data.
Currently I am trying deploy the application locally (with docker) on the google cloud shell. But I am totally not sure if this is the right way and I am also really struggling to get it on GCS.
Could someone help me please?
Awesome to hear about your project @lennartsteenberghs. Apologies for the late reply. Is there a reason you don't want to use the web app? https://www.trashai.org/
How would you want to interact with an API?
Hi, I know I wasn't the one who initiated the request but I do have a relevant question, me and my friends wanted to use an API of this model for an app we are building (open-source), and we have a paged dedicated to trash classification. Now I was looking for a specific endpoint that I can upload and retrieve an image through the app because the web app does not help with this task. Let me know if you have anything currently that could help me.
Thank you!
Would be happy to build it out and add it, would need access to the model itself, I do believe I already have the code for mapping the classifications. I looked through GitHub and Docker, but I could not find the model itself. If you have a directory or a link, I could start building it.