Eye movement track identification method and system
A trajectory and eye movement technology, applied in neural learning methods, eye testing equipment, character and pattern recognition, etc., can solve problems such as time-consuming, labor-intensive, highly subjective, and inability to receive treatment as soon as possible.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0050] Embodiment 1 of the present invention provides an eye movement trajectory identification system, which includes:
[0051] An acquisition module, configured to acquire the eye track to be identified;
[0052] The identification module is used to use the trained identification model to process the acquired eye movement trajectory to be identified, and obtain the identification result that the eye movement track to be identified is normal or abnormal; wherein, the trained identification model Obtained by training with a training set, the training set includes original pictures and corresponding normal children's visual salience pictures and autistic children's visual salience pictures, and the original pictures are respectively obtained by multiple normal children and multiple autistic children Browse and mark the location information of the fixation point and the duration of the fixation point.
[0053] In this embodiment 1, the above-mentioned system is used to realize ...
Embodiment 2
[0088] like figure 1 As shown, in the present embodiment 2, a kind of children's autism detection method based on eye movement track identification is provided, in this method, network is trained to obtain the discrimination model of eye track, and the network of training mainly comprises by 3 volumes A convolutional deep network composed of product layers, 2 pooling layers and a fully connected layer, and a long short-term memory network composed of a double-layer long short-term memory network layer, a fully connected layer and a Sigmoid layer. The model includes two processes of training and recognition:
[0089] Both the training process and the recognition process use the Saliency4ASD data set, such as figure 2 as shown, figure 2 (a) is the original picture of the data set, and the subject recorded the eye movement trajectory after viewing the original picture, that is, the coordinates and time information of the gaze point, figure 2 (b) for figure 2 (a) The corre...
Embodiment 3
[0133] Embodiment 3 of the present invention provides a non-transitory computer-readable storage medium, the non-transitory computer-readable storage medium includes instructions for executing an eye movement trajectory identification method, and the method includes:
[0134] Obtain the eye track to be identified;
[0135] Utilize the trained identification model to process the obtained eye movement track to be identified, and obtain the identification result that the eye track to be identified is normal or abnormal; wherein, the trained identification model is trained using a training set Obtained, the training set includes original pictures and corresponding normal children's visual salience pictures and autistic children's visual salience pictures, and the original pictures are browsed by a plurality of normal children and a plurality of autistic children respectively. Position information and gaze point duration are marked.
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More