Dimensionality's blessing: Clustering images by underlying distribution

• Main Authors: LIN, Wen-yan, LAI, Jian-Huang, LIU, Siying, MATSUSHITA, Yasuyuki
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2018
• Subjects: Computer and Systems Architecture; Graphics and Human Computer Interfaces
• Online Access: https://ink.library.smu.edu.sg/sis_research/4810; https://ink.library.smu.edu.sg/context/sis_research/article/5813/viewcontent/Lin_Dimensionalitys_Blessing_Clustering_CVPR_2018_paper.pdf
• Institution: Singapore Management University

Many high dimensional vector distances tend to a constant. This is typically considered a negative “contrastloss” phenomenon that hinders clustering and other machine learning techniques. We reinterpret “contrast-loss” as a blessing. Re-deriving “contrast-loss” using the law of large numbers, we sho...