Jeongin Seo

Bio: Jeongin Seo is an academic researcher from Kyungpook National University. The author has contributed to research in topics: Three-dimensional face recognition & Facial recognition system. The author has an hindex of 3, co-authored 10 publications receiving 56 citations.

Object Recognition in Very Low Resolution Images Using Deep Collaborative Learning

Abstract: Although recent studies on object recognition using deep neural networks have reported remarkable performance, they have usually assumed that adequate object size and image resolution are available, which may not be guaranteed in real applications. This paper proposes a framework for recognizing objects in very low resolution images through the collaborative learning of two deep neural networks: image enhancement network and object recognition network. The proposed image enhancement network attempts to enhance extremely low resolution images into sharper and more informative images with the use of collaborative learning signals from the object recognition network. The object recognition network with trained weights for high resolution images actively participates in the learning of the image enhancement network. It also utilizes the output from the image enhancement network as augmented learning data to boost its recognition performance on very low resolution objects. Through experiments on various low resolution image benchmark datasets, we verified that the proposed method can improve the image reconstruction and classification performance.

A robust face recognition through statistical learning of local features

Abstract: Among various signals that can be obtained from humans, facial image is one of the hottest topics in the field of pattern recognition and machine learning due to its diverse variations. In order to deal with the variations such as illuminations, expressions, poses, and occlusions, it is important to find a discriminative feature which can keep core information of original images as well as can be robust to the undesirable variations. In the present work, we try to develop a face recognition method which is robust to local variations through statistical learning of local features. Like conventional local approaches, the proposed method represents an image as a set of local feature descriptors. The local feature descriptors are then treated as a random samples, and we estimate the probability density of each local features representing each local area of facial images. In the classification stage, the estimated probability density is used for defining a weighted distance measure between two images. Through computational experiments on benchmark data sets, we show that the proposed method is more robust to local variations than the conventional methods using statistical features or local features.

Multi-Attribute Recognition of Facial Images Considering Exclusive and Correlated Relationship Among Attributes

Abstract: Multi-attribute recognition is one of the main topics attaining much attention in the pattern recognition field these days. The conventional approaches to multi-attribute recognition has mainly focused on developing an individual classifier for each attribute. However, due to rapid growth of deep learning techniques, multi-attribute recognition using multi-task learning enables the simultaneous recognition of more than two relevant recognition tasks through a single network. A number of studies on multi-task learning have shown that it is effective in improving recognition performance for all tasks when related tasks are learned together. However, since there are no specific criteria for determining the relationship among attributes, it is difficult and confusing to choose a good combination of tasks that have a positive impact on recognition performance. As one way to solve this problem, we propose a multi-attribute recognition method based on the novel output representations of a deep learning network which automatically learns the exclusive and joint relationship among attribute recognition tasks. We apply our proposed method to multi-attribute recognition of facial images, and confirm the effectiveness through experiments on a benchmark database.

Matrix correlation distance for 2D image classification

Abstract: In the field of visual information processing, there have been active studies on the efficient representation of visual data, such as local feature descriptors and tensor subspace analysis. Though these methods give a representation using matrix features, current methods for classification are mainly designed for 1D vector data, which may lead to loss of information included in 2D matrix data. To solve the problem, we propose a matrix correlation distance for 2D image data by extending the correlation distance for random vectors. Through a number of computational experiments on image data with various representations, we compare the performance of the proposed measure with conventional distances.

Application Research of Improved YOLO V3 Algorithm in PCB Electronic Component Detection

Abstract: Target detection of electronic components on PCB (Printed circuit board) based on vision is the core technology for 3C (Computer, Communication and Consumer Electronics) manufacturing companies to achieve quality control and intelligent assembly of robots. However, the number of electronic components on PCB is large, and the shape is different. At present, the accuracy of the algorithm for detecting all electronic components is not high. This paper proposes an improved algorithm based on YOLO (you only look once) V3 (Version 3), which uses a real PCB picture and a virtual PCB picture with synthesized data as a joint training dataset, which greatly increases the recognizability of training electronic components and provides the greatest possibility for data enhancement. After analyzing the feature distribution of the five dimensionality-reduced output layers of Darknet-53 and the size distribution of the detection target, it is proposed to adjust the original three YOLO output layers to four YOLO output layers and generate 12 anchor boxes for electronic component detection. The experimental results show that the mean average precision (mAP) of the improved YOLO V3 algorithm can achieve 93.07%.

Sensing system of environmental perception technologies for driverless vehicle: A review of state of the art and challenges

Abstract: Environmental perception technology is the guarantee of the safety of driverless vehicles. At present, there are a lot of researches and reviews on environmental perception, aiming to realize unmanned driving while ensuring the safety of human life. However, the technology is facing new challenges in the new era. This review paper attempts to systematically summarize environment perception technology and discuss the new challenges currently faced. To this end, we first summarized the advantages, disadvantages and applicable occasions of several commonly used sensing methods to provide a clear selection guide. The new challenges faced by environmental perception technology are discussed from three aspects: technology, external environment and applications. Finally, the article also points out the future development trends and efforts of environmental perception technology.