Unlock instant, AI-driven research and patent intelligence for your innovation.

A Gaze Trajectory Prediction Method Based on Object Detection for Category Search Task

A target detection and trajectory prediction technology, applied in the field of computer vision, can solve problems such as unusable research, complex background images, etc., to avoid direct training, reduce dependencies, and improve interpretability.

Active Publication Date: 2021-10-19
XI AN JIAOTONG UNIV
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This method obtains the characteristics of the target category through bag of words and clustering. It requires the background of the image to be searched to be simple, which is convenient for converting it into the required saliency map. However, the actual background image is more complex, so it cannot be used in actual research. middle

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A Gaze Trajectory Prediction Method Based on Object Detection for Category Search Task
  • A Gaze Trajectory Prediction Method Based on Object Detection for Category Search Task
  • A Gaze Trajectory Prediction Method Based on Object Detection for Category Search Task

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0031] Embodiments of the present invention are as follows figure 1 As shown in the figure, the main steps contained therein are as follows:

[0032] Step 1: Blur the gaze point;

[0033] Step 2: Feature map extraction;

[0034] Step 3: Predict the gaze point;

[0035] Step 4: Determine whether to stop;

[0036] Step 5: suppress graph update;

[0037] The specific implementation process of step 1 is as follows:

[0038] This method uses a multi-resolution pyramid to create a resolution map that indicates the degree of low-pass filtering applied to each image point relative to its current fixed distance, providing a basis for subsequent processing, which is divided into the following Four steps, corresponding to figure 2 :

[0039] 1) Separate the RGB channels and encode them as 1, 2, and 3 respectively;

[0040] 2) Create a resolution map;

[0041] 3) Apply the encoding to the resolution map;

[0042] 4) Combine the three channels to obtain a blurred image;

[0043...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a gaze track prediction method based on target detection for category search tasks, which belongs to the field of computer vision; mainly including the following steps: 1. Fuzzy the picture according to the characteristics of human eyes; 2. According to the category of the target to be searched , convert the blurred picture into a saliency map corresponding to the image processed by the human eye; 3. According to the saliency map and the suppression map, predict the current picture and generate the current predicted gaze position; 4. If the gaze position does not reach the target point or has reached the maximum number of fixation times, you can end this fixation point prediction, otherwise, suppress the current fixation position, update the suppression point map, and repeat steps 1 to 3; 5. Connect the generated fixation positions in sequence according to their generation order , the gaze trajectory can be obtained. This method can not only adapt to the images to be searched with complex backgrounds, but also consider the actual physiological constraints of the human eye.

Description

technical field [0001] The invention belongs to the field of computer vision, and in particular relates to a category search task-oriented gaze trajectory prediction method based on faster-RCNN. A superior colliculus model to predict gaze trajectories under category search tasks. Background technique [0002] Analyzing the human eye is important for understanding human behavior and cognitive patterns. The way the human eye processes information is very complex, and many processing processes are not clearly explained today. For the human eye, the expressions we can clearly know are saccade, eye movement trajectory, etc. At present, the task background of gaze trajectory prediction mainly focuses on two directions, namely free browsing behavior without task-driven and task-driven search behavior. The former promotes the research of saliency map methods, and the development of the latter promotes People's research on the physiological mechanism of the human eye and the resear...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06F3/01G06K9/00G06N3/04
CPCG06F3/013G06V40/197G06V40/193G06V40/19G06V2201/07G06N3/045
Inventor 张雪涛卢美祺吴镜涵林玉萍郑南宁
Owner XI AN JIAOTONG UNIV