Congratulations to Kaitai Zhang for passing his defense on May 19, 2021. His Ph.D. thesis is entitled “Data-Driven Image Analysis, Modeling, Synthesis and Anomaly Localization Techniques”. Here we invite Kaitai to share a brief introduction of his thesis and some words he would like to say at the end of the Ph.D. study journey.

1) Abstract of Thesis

Emerging Deep learning and machine learning techniques have brought impressive improvements for numerous topics in image processing and computer vision fields. In this thesis, we introduce our research on Data-Driven Image Analysis, Modeling, Synthesis and Anomaly Localization Techniques: 1) image anomaly detection and localization; 2) texture analysis, modeling and synthesis.

For the first part, we will focus on image anomaly detection and localization tasks. Image anomaly detection is a binary classification problem to determine whether an input contains an anomaly, and image anomaly localization is to get pixel-precise segmentation of regions that appear anomalous. Detecting and localizing anomalies is a critical and long-standing problem in image processing and computer vision, and has applications in many real-world scenarios like medical image diagnosis and automated manufacturing inspection. In this talk, I will introduce two of our recent works, PEDENet and AnomalyHop. PEDENet is a neural network-based framework that jointly learns image local feature and density estimation model. AnomalyHop employs successive subspace learning (SSL) framework, and utilizes various Gaussian Descriptors to learn normality feature distributions. Both of them achieve state-of-the-art performance on MVTec AD dataset, and provide either smaller model size or faster inference speed.

In the second part, our previous works in texture analysis, modeling and synthesis will be reviewed. For dynamic texture synthesis, two effective techniques will be proposed and proved effective. The enhanced model could encode coherence of local features as well as the correlation between local feature and its neighbors, and also capture more complicated motion in the time domain. For unsupervised texture segmentation, an effective textural feature extraction method will be introduced, which offers the state-of-the-art performance. Features are learned from data in an unsupervised manner. They encode local features as well as contrast information. We will also talk about an effective hierarchical spatial-spectral correlation (HSSC) method for texture analysis and classification.

2) Ph.D. experience:

I would like to thank Professor Kuo for all his support and kindness to me in the past a few years. Professor Kuo is a true scholar who lives by example and motivates me during my entire PhD journey.  Besides inspiring guidance on my research, Professor Kuo also teach me his lifelong wisdom, like independent thinking, self-discipline, and his dedicated attitude to research. All of them are something I need to learn throughout my life. I also want to thank all my MCL peers as well for their insight discussions and encouragement. The life-long friendship with them is one of the most valuable parts of my PhD journey. I would like to wish the best to all the people in our lab in their future study!