3.4.3. Pet Stories

Pet detection based on LLM. When a pet appears in front of the camera, it will send a notification to alert the user.

Agent Logic

First, Pet Stories inputs the user id, the device used, and the image that needs to be detected through the Start node. The image is stored by the Media node and then checked by the Pet Detection Extension node to see if a pet is present and to identify the type of pet.

Then, the Program node extracts the Boolean value of hasPet from the Pet Detection node (True indicates a pet is present in the image, False indicates there is none) and uses the Condition node to evaluate the value of hasPet.

If a pet is present in the image, LLM is used to describe the pet's behavior in the image. The information is uploaded to the cloud for easy user access, and finally, the pet's current behavior and image are pushed to the user through the Message push node.

Agent Node Usage

After duplicating Pet Stories, you can test run and publish Pet Stories, allowing it to be used as a node in the Agent&Agent Node of other projects. By inputting the media parameters imported through the Start node (without needing to be stored by the Media node) and the session, the node's input data requirements can be met.