Receiving and saving inference feedback in the datastore

[Francois-Werbrouck]

Context:
Following ai-cfia/nachet-frontend#129 and ai-cfia/nachet-backend#87 the datastore needs to be able to register the feedback from the user on an inference. Once a picture is confirmed, we need to create the picture_seed entity

Type of feedback:

• Perfect
• Not a seed
• Corrected seed / Corrected box metadata

Acceptance Criteria

1. The datastore has high level function to interact with the BE to receive the feedback result
2. The datastore metadata module can handle the feedback
3. The datastore updates the appropriated rows in the database
4. A picture_seed entity is created once a picture seed is confirmed
5. Each object metadata is updated if necessary (position, color, etc...)
6. A corresponding test suite

Todo:

• Implement the a perfect feedback function #29
• Implement the not a seed feedback function #39
• Implement the Correction feedback function #40
• Create a trigger in the DB once an inference is verified to insert the new picture_seed #41

[Francois-Werbrouck]

We are currently in a testing stage with our users