slack-classify-image

Example: Classify an Image posted in Slack with Leonardo

A function is triggered by an image upload to a Slack channel. The function receives an event about the image upload. It then invokes the SAP Leonardo image classification API with the image and waits for the response. The response itself contains a prediction about the possible content of the image. The function then posts the prediction to the channel.

Requirements

• A Slack application, for example within a test workspace, is required. The Slack application should define a bot user and be subscribed to the file_shared bot event. Bot events are used, because image classification should be limited to channels where the bot user is invited.

Deployment

First, create a deployment file to provide credentials. Run inside the project directory:

faas-sdk init-values -y values.yaml

Update the generated file:

• An api key for the image classification can be obtained at Leonardo image classification overview.
• Also specify the Slack bot user token.

And finally, deploy the project :

 xfsrt-cli faas project deploy -y ./deploy/values.yaml -v

Configure Slack Endpoint

Go to the Configuration UI of your Slack application, navigate to Event Subscriptions and toggle Enable Events. Configure the HTTP trigger URL as the corresponding slack bot event request endpoint URL. The HTTP trigger URL can be retrieved from:

xfsrt-cli faas project get slack

The output contains the trigger URL.

Test

Install the Slack application to your workspace. In Slack, create a channel. Invite the bot to your channel. Click on the attachment upload button inside the channel, select an image of your choice and upload it. As a response, you should get an SAP Leonardo powered probability estimation of what is contained in the image.

License

Copyright (c) 2020 SAP SE or an SAP affiliate company. All rights reserved. This file is licensed under the Apache Software License, v. 2 except as noted otherwise in the LICENSE file.