報(bào)告題目：Tensor Factorization with Total Variation for Low-Rank Tensor Completion in Imaging Data

報(bào)告時(shí)間：2020年6月29日星期一 8:30

報(bào)告方式：騰訊會(huì)議（ID：916 460 699）

報(bào)告內(nèi)容：In this talk, Iwill focus on tensor factorization for low-rank tensor completion in imaging data.Due to the underlying redundancy of real-world imaging data, the low-tubal-rank tensor factorization (the tensor-tensor product of two factor tensors) can be used to approximate such tensor tensors very well.Motivated by the spatial/temporal smoothness of factor tensors in real-world imaging data, a low-tubal-rank tensor factorization model with a hybrid regularization combining total variation and Tikhonov regularization for low-rank tensor completion problem is developed.A proximal alternating minimization (PAM) algorithm is employed to tackle the corresponding minimization problem. A global convergence the PAM algorithm is established. Numerical results on color images, color videos, and multi-spectral images (MSIs)are reported to show the performance of the proposed method.

嘉賓簡(jiǎn)介

林學(xué)磊，香港浸會(huì)大學(xué)博士。2014年在寧夏大學(xué)獲得信息與計(jì)算科學(xué)專(zhuān)業(yè)理學(xué)學(xué)士學(xué)位；2017年在澳門(mén)大學(xué)獲得數(shù)學(xué)專(zhuān)業(yè)理學(xué)碩士學(xué)位；2017年至今在香港浸會(huì)大學(xué)數(shù)學(xué)系攻讀博士學(xué)位。主要從事數(shù)值線(xiàn)性代數(shù)方面的研究，包括偏微分方程數(shù)值解、結(jié)構(gòu)線(xiàn)性系統(tǒng)的快速迭代法、張量計(jì)算在數(shù)字圖像處理方面的應(yīng)用，已在J. Comput. Phys.，SIAM J. Matrix Anal. Appl.，J. Sci. Comput.，BIT. Numerical Mathematics，SIAM J. Numer. Anal.，Comput. Math. Appl.，J. Math. Imaging Vision等刊物以第一作者身份發(fā)表論文10余篇。曾在北京清華大學(xué)召開(kāi)的“第八屆世界華人數(shù)學(xué)家大會(huì)”上，獲2019年“新世界數(shù)學(xué)獎(jiǎng)”的優(yōu)秀碩士論文銀獎(jiǎng)，是澳門(mén)首次獲得該獎(jiǎng)項(xiàng)，獲“第十四屆東亞工業(yè)與應(yīng)用數(shù)學(xué)學(xué)會(huì)年會(huì)”優(yōu)秀學(xué)生論文獎(jiǎng)二等獎(jiǎng)，獲香港政府博士獎(jiǎng)學(xué)金，獲澳門(mén)研究生科技研發(fā)獎(jiǎng)。

                                        寧夏大學(xué)數(shù)學(xué)統(tǒng)計(jì)學(xué)院 寧夏師范學(xué)院數(shù)學(xué)與計(jì)算機(jī)學(xué)院 聯(lián)合承辦

2020年6月28日