Overview

Estimate age and gender using Grip's demographics models.

Inputs & Outputs

Input media: Raw Video.
Output media: Raw Video.
Output metadata: objects[].attributes (age and gender).

Pipeline usage

Place a face detector and Grip demographics model upstream. This node only evaluates objects whose label is face, and expects the corresponding object tensor to contain a 12-value age/gender classification layer. A tracker is strongly recommended because estimates are cached by object ID between model inference frames.

For each recognized face, the node appends two attributes to the object:

A numeric age estimate with class ID 999.
male or female with class ID 998.

max_object_lifespan controls how long a cached estimate is reused after its object ID disappears. Use gender_bias_adjustment only after evaluating the model on representative camera placement and lighting; positive values bias toward female, while negative values bias toward male.

Properties

Property	Description	Type	Default	Required
`max_object_lifespan`	Max. object lifespan. Unit: seconds.	`float`	`60`	No
`gender_bias_adjustment`	Adjust for gender skew introduced by environmental factors. Negative values adjust towards male, positive values adjust towards female. Range: minimum `-0.5`, maximum `0.5`. Step: `0.05`.	`float`	`0.15`	No

Output Metadata

The fields below are declared by this node's metadata schema; the JSON values are representative examples.

Path	Type	Description
`objects`	`array`	Value of objects.
`objects[].attributes`	`array`	Value of attributes.
`objects[].attributes[].class_id`	`integer`	Value of class id.
`objects[].attributes[].label`	`string`	Value of label.
`objects[].attributes[].probability`	`number`	Value of probability.
`objects[].class_id`	`integer`	Value of class id.
`objects[].id`	`integer`	Value of id.
`objects[].label`	`string`	Value of label.
`objects[].probability`	`number`	Value of probability.
`objects[].rect.height`	`integer`	Value of height.
`objects[].rect.left`	`integer`	Value of left.
`objects[].rect.top`	`integer`	Value of top.
`objects[].rect.width`	`integer`	Value of width.

JSON example

{
  "objects": [
    {
      "attributes": [
        {
          "class_id": 999,
          "label": "20",
          "probability": 1.0
        },
        {
          "class_id": 998,
          "label": "female",
          "probability": 1.0
        }
      ],
      "class_id": 1,
      "id": 1,
      "label": "face",
      "probability": 0.9,
      "rect": {
        "height": 40,
        "left": 20,
        "top": 20,
        "width": 40
      }
    },
    {
      "attributes": [],
      "class_id": 1,
      "id": 9,
      "label": "person",
      "probability": 0.9,
      "rect": {
        "height": 120,
        "left": 10,
        "top": 10,
        "width": 60
      }
    },
    {
      "attributes": [
        {
          "class_id": 999,
          "label": "66",
          "probability": 1.0
        },
        {
          "class_id": 998,
          "label": "male",
          "probability": 1.0
        }
      ],
      "class_id": 1,
      "id": 2,
      "label": "face",
      "probability": 0.9,
      "rect": {
        "height": 40,
        "left": 20,
        "top": 20,
        "width": 40
      }
    }
  ]
}

Object labels and attributes

Object labels/classes added: None.
Object attribute labels/classes added: Estimated age values use class 999; estimated gender values use class 998.