BaselHack_Product_Search

Product Search project for BaselHack 2021

Important Notes

• Create a Google account with credit card / billing enabled.
• Follow the instructions from the quick start guide of the Google Vision API Product Search: [https://cloud.google.com/vision/product-search/docs/quickstart]
• Upload the pictures into a Bucket on Google storage.
• Upload the CSV describing the pictures into a Bucket on Google storage. ** make sure the csv have enough rows, use empty rows
• Create a service principal, download the token, set the environment variable.
• Make sure this service principal has access to the bucket with the CSV and the pictures.
• Run the request to start the training. It will return a request to ask for the status.
• It seems to be required to put all your products into one 'ProductSet' because you have to provide it in the request.
• As soon as the training is finished, you can ask the API to search pictures. The results are acceptable.
• When you use one of the training pictures, it will most probably find the right one, but not with a match of 100%. Therefore we assume that Google uses the product category (e.g. packagedgoods-v1) to recogize the object itself (e.g. a package of chips) and then knows how to deal with angle and orientation and then maps the product itself. ** hits under 0.5 score can be dismissed
• If something does not work, try it again after 30 minutes - only then Google updates the data. ** changes in the product data is only loaded in 30 minutes bulks

Details

Technology stack

• Angular
• node.js
• Docker
• Vision API Product Search from Google Cloud

Test deployment

Proxy

[https://levelup.gitconnected.com/fixing-cors-errors-with-angular-cli-proxy-e5e0ef143f85]

npm install cors

const cors = require('cors');
app.use(cors({
    origin: ['https://www.section.io', 'https://www.google.com/']
}));