SeeSo SDK is a library that calculates where the user is looking at on the screen in real time using RGB image from the camera.
All computation is done locally on the device.
What SeeSo SDK provides
We define Gaze as the position where the user is looking at.
SeeSo SDK provides Gaze in 2D screen coordinates (x, y).
SeeSo SDK classifies two types of Eye Movements:
- Fixation: When the user's gaze is focused on a specific position
- Saccade: When the user's gaze is moving from one point to another
What to prepare
Before using SeeSo SDK, you need to [generate a license key](https://manage.seeso.io/#/console/license-keys\).
Network connected environment
Because SeeSo SDK requires authentication, network connected environment is required.
SeeSo SDK requires image of the user from a camera as its input data.
For mobile devices, this can be frontal-camera. For laptops and desktops, if build-in camera is not available, extra hardware is required.
How SeeSo SDK works
Step 1: Authentication
SeeSo SDK validates the
license key and check its match to the
If the authentication is successful, the SeeSo Server will issue a session and the SeeSo SDK will construct a GazeTracker instance.
Step 2: Frontal Face Image Capture
GazeTracker instance captures frontal face image to calculate 2D gaze points.
Step 3: Gaze & Eye Movement Calculation
GazeTracker instance executes pre-processing, image processing, and deep-learning computation on the captured image.
The calculated outputs (Gaze and Eye Movement) will be passed to your application through callbacks.
- Match to application information is only checked when a production key is used.↩