Perception Encoder Embedding Model

Generate an embedding of an image or string.

Use the Meta Perception Encoder model to create semantic embeddings of text and images.

This block accepts an image or string and returns an embedding. The embedding can be used to compare similarity between different images or between images and text.

Type identifier

Use the following identifier in step "type" field: roboflow_core/perception_encoder@v1 to add the block as a step in your workflow.

Properties

Name	Type	Description	Refs
name	str	Unique name of step in workflows.	❌
data	str	The string or image to generate an embedding for..	✅
version	str	Variant of Perception Encoder model.	✅

The Refs column marks possibility to parametrise the property with dynamic values available in workflow runtime. See Bindings for more info.

Runtime compatibility

hard - runtime self_hosted_cpu; execution local : Requires a GPU; run_locally() loads a model that needs CUDA.

Input and Output Bindings

The available connections depend on its binding kinds. Check what binding kinds Perception Encoder Embedding Model in version v1 has.

Input and output bindings

• input

  • data (Union[image, string]): The string or image to generate an embedding for..

  • version (string): Variant of Perception Encoder model.

• output

  • embedding (embedding): A list of floating point numbers representing a vector embedding..

Example JSON definition

{
	    "name": "<your_step_name_here>",
	    "type": "roboflow_core/perception_encoder@v1",
	    "data": "$inputs.image",
	    "version": "PE-Core-B16-224"
	}