Head keypoints#

Overview#

The ground truth exposed in this modality is a set of 81 keypoints that define the overall shape of the head, using a standard developed by Datagen.

This modality consists of the following files:

Relevant file	Location
head_key_points.json	key_points folder
all_key_points.json	key_points folder

head_key_points.json#

This file contains information on the structure of the subject’s head. It has the following structure:

"version": "2.0.0",
"head": {
   "top": {
        // 12 keypoints defining the structure of the top of the head
   }
   "left": {
        // 24 keypoints defining the structure of the left side of the head
   }
   "back": {
        // 21 keypoints defining the structure of the back of the head
   }
   "right": {
        // 24 keypoints defining the structure of the right side of the head
   }
}

Note

For keypoints that define the structure of the front of the head, i.e., the face, see the iBug and MediaPipe facial keypoint modalities.

Objects and fields:

version: String. Version tracking for this file. Whenever you access this file in a datapoint, make sure to check that the version matches what you expect it to be; otherwise its format and fields may not be recognized.
top: Object. This object contains 12 objects, each identifying a specific keypoint on the top of the head. The keypoint objects are named “1” through “12”, and each one has the same format:
```
"3": {
   "global_3d": {
      "x": 0.705421507358551,
      "y": 0.18709957599639893,
      "z": 0.044457197189331055
   },
   "pixel_2d": {
      "x": -26.041956551860782,
      "y": 75.9609971318042
   },
   "is_visible": "false"
```
- global_3d: Object. Contains a set of three Floats giving the location of the keypoint in global coordinates: “x”, “y”, and “z”. See About our coordinate systems for details.
- pixel_2d: Object. Contains a set of two Ints giving the x and y coordinates of the keypoint in the rendered image itself. See About our coordinate systems for details.
- is_visible: Boolean. Indicates whether the keypoint is visible in the images produced by this camera. The value of this field is false if the keypoint is outside the frame or is on the other side of the body from the camera. Otherwise the value is true, even if there is another object between the keypoint and the camera.
The 12 keypoints on the top of the head.#
left: Object. This object has the same structure as “top” above, except it identifies 24 keypoints on the left side of the head.

The 24 keypoints on the left side of the head.#
back: Object. This object has the same structure as “top” above, except it identifies 21 keypoints on the back of the head.

The 21 keypoints on the back of the head.#
right: Object. This object has the same structure as “top” above, except it identifies 24 keypoints on the right side of the head.

The 24 keypoints on the right side of the head.#

Using this ground truth, you can train your model to recognize the shape and structure of the human head.

all_key_points.json#

For convenience, this file includes the same set of keypoints as head_key_points.json above, located at this position in the hierarchy:

{
    "head": {
        "top": {
             // 12 keypoints defining the structure of the top of the head
        }
        "left": {
             // 24 keypoints defining the structure of the left side of the head
        }
        "back": {
             // 21 keypoints defining the structure of the back of the head
        }
        "right": {
             // 24 keypoints defining the structure of the right side of the head
        }
    }
}

For details on the structure of the keypoint objects, see head_key_points.json above.