Saliency detection is an effective way to acquire potential regions of interest that may attract human eyes, its numerous applications range from object detection and recognition, image compression, video summarization, to content-based image editing and image retrieval. Towards better grouping of objects and background, a method based on Normalized graph cut (Ncut) is proposed for saliency detection. Since the Ncut partitions a graph in a normalized energy minimization fashion, resulting eigenvectors of the Ncut contain good cluster information that may group visual contents. Motivated by this, the proposed method directly induces saliency maps via eigenvectors of the Ncut, contributing to accurate saliency estimation of visual clusters.
Given a few labelled samples, semi-supervised learning generally performs much better than supervised learning, semi-supervised learning algorithms are more robust to noise. Label Prediction via Deformed Graph Laplacian for Semi-supervised Learning is presented. A novel curriculum learning approach, dubbed multi-modal curriculum learning, to optimize the quality of semi-supervised image classification is proposed.
Speaker(s): Jie Yang,
Location:
Room: ASB 10940 (SFU's Big Data Visualization Lab)
Bldg: Applied Sciences Building
Simon Fraser University
8888 University Drive
Burnaby, British Columbia
V5A 1S6