Changqun Xia

Title: Innovations of Changqun Xia in Image Segmentation

Introduction

Changqun Xia is a prominent inventor based in Beijing, China. He has made significant contributions to the field of image processing, particularly in the area of salient object segmentation. With a total of four patents to his name, Xia's work is recognized for its innovative approaches and practical applications.

Latest Patents

Changqun Xia's latest patents include an "Image Salient Object Segmentation Method and Apparatus Based on Reciprocal Attention Between Foreground and Background." This method involves obtaining a feature map from a training image using a convolutional neural backbone network. It further includes obtaining initial feature responses for both foreground and background, updating these responses with a reciprocal attention weight matrix, and training a convolutional neural network model for segmentation.

Another notable patent is the "Method for Salient Object Segmentation of Image by Aggregating Multi-Linear Exemplar Regressors." This method analyzes visual attributes of salient and non-salient objects, constructs a quadratic optimization problem, and generates a final foreground probability map. It also involves training linear exemplar regressors to characterize saliency patterns, ultimately forming a robust image segmentation model.

Career Highlights

Changqun Xia is affiliated with Beihang University, where he continues to advance research in image processing and computer vision. His work has garnered attention for its innovative methodologies and practical implications in various applications.

Collaborations

Some of his notable coworkers include Jia Qi Li and Qinping Zhao, who have collaborated with him on various research projects and patents.

Conclusion

Changqun Xia's contributions to image segmentation through his innovative patents highlight his expertise and dedication to advancing technology in this field. His work continues to influence the landscape of image processing and computer vision.