IEEE CIS    |    CIM    |    TNNLS    |    TEVC    |    TFS    |    TCDS    |    TETCI    |    TG
IEEE CIS Newsletter, Issue 83, December 2019

Research Frontier

Object Detection With Deep Learning: A Review

Due to object detection's close relationship with video analysis and image understanding, it has attracted much research attention in recent years. Traditional object detection methods are built on handcrafted features and shallow trainable architectures. Their performance easily stagnates by constructing complex ensembles that combine multiple low-level image features with high-level context from object detectors and scene classifiers. With the rapid development in deep learning, more powerful tools, which are able to learn semantic, high-level, deeper features, are introduced to address the problems existing in traditional architectures. These models behave differently in network architecture, training strategy, and optimization function. In this paper, we provide a review of deep learning-based object detection frameworks. Our review begins with a brief introduction on the history of deep learning and its representative tool, namely, the convolutional neural network. Then, we focus on typical generic object detection architectures along with some modifications and useful tricks to improve detection performance further. As distinct specific detection tasks exhibit different characteristics, we also briefly survey several specific tasks, including salient object detection, face detection, and pedestrian detection. Experimental analyses are also provided to compare various methods and draw some meaningful conclusions. Finally, several promising directions and tasks are provided to serve as guidelines for future work in both object detection and relevant neural network-based learning systems.

IEEE Transactions on Neural Networks and Learning Systems, Nov. 2019

Fuzzy Rough Set Based Feature Selection for Large-Scale Hierarchical Classification

The classification of high-dimensional tasks remains a significant challenge for machine learning algorithms. Feature selection is considered to be an indispensable preprocessing step in high-dimensional data classification. In the era of big data, there may be hundreds of class labels, and the hierarchical structure of the classes is often available. This structure is helpful in feature selection and classifier training. However, most current techniques do not consider the hierarchical structure. In this paper, we design a feature selection strategy for hierarchical classification based on fuzzy rough sets. First, a fuzzy rough set model for hierarchical structures is developed to compute the lower and upper approximations of classes organized with a class hierarchy. This model is distinguished from existing techniques by the hierarchical class structure. A hierarchical feature selection problem is then defined based on the model. The new model is more practical than existing feature selection approaches, as many real-world tasks are naturally cast in terms of hierarchical classification. A feature selection algorithm based on sibling nodes is proposed, and this is shown to be more efficient and more versatile than flat feature selection. Compared with the flat feature selection algorithm, the computational load of the proposed algorithm is reduced from 98.0% to 6.5%, while the classification performance is improved on the SAIAPR dataset. The related experiments also demonstrate the effectiveness of the hierarchical algorithm.

IEEE Transactions on Fuzzy Systems, Oct. 2019

Educational Activities

2019 IEEE CIS Summer School on Computational Intelligence Techniques for High School Students

The wave of AI enthusiasm triggers broad attention to “intelligent technology.” Many recent highly-attentive AI techniques were in CIS main technology fields. For a long time now, CIS served as the Society leading the advance of theory and the development of systems. How to make the contributions of CIS knowne, and bring the impact of CIS to broader society and down to the young generation are very important in this moment.

Under this consideration, the main objective of this summer school is to bring CIS techniques to talented high school students. The program aims to provide a better understanding of the CIS techniques and potential applications before students enter college, and therefore generate interest in this area during collegiate study.

Another purpose of this summer school is to trigger more female high school students to study in CIS-related techniques. As we know, gender imbalance in engineering is a serious issue. In order to interest more female high school students on CIS-related techniques, we particularly invited, and gave higher priority to them.

The program was organized on 31 July - 2 August 2019, a three-day summer school on computational intelligence techniques held on National Cheng Kung University, Tainan, Taiwan. The program contained six lectures and four hand-on experiments. The topics of the six lectures are: Supervised Learning VS Unsupervised Learning, Decision Trees and Random Forest, Deep Learning Networks for Image Analysis, Multilingual and Cross-lingual Information Analysis, Introduction to Evolutionary Computing, and Evolutionary Computing in Music Arts and Creativity. In addition to the lectures, we also designed four hand-on experiments, for which eight graduate students were allocated to serve as assistants during the hand-on experiments. During the hand-on experiments, the students had a chance to write fragments of programs and experienced the application of Taiwanese speech recognition technology by using Raspberry Pi as platform. Students were amazed by the computational intelligence techniques that can be used in an application of understanding Taiwanese dialog, which is a language that most of the current students are not familiar with. They were also intrigued by the capability of Evolutionary Computing and asked for the music created by the EC for their collection.

Nvidia instructor was also invited to guide students on a hands-on experiment on a Generative Adversarial Network (GAN) and its application for creating a variety of cat models. Another demo of GAN transforming a photo provided by students into Picasso-style was also presented. The students were amazed by the ability of the algorithm.

There were 31 male and 21 female students registered. In the end, 28 male and 13 female students attended. The students were from all over the country and stayed in the university dormitory. They had a very positive response to the curriculum and expressed their appreciation for having such an opportunity.

The organizers would particularly thank Gary Yen travelling from the USA, and Chuan-Kang Ting from northern Taiwan for offering courses on Evolutionary Computing.

Call for Applications: Global Scientist Interdisciplinary Forum 2020 (15 Dec)

The Global Scientist Interdisciplinary Forum at Southern University of Science and Technology (SUSTech) is an important conference for talent recruitment. It aims to provide a platform for young scholars at home and abroad to network and exchange their research ideas, and to find out the ambitious goals of SUSTech becoming a world-class research university. It also provides a unique forum for the participants to interact with existing faculty members of SUSTech in Shenzhen, China.

Forum Schedule
Application Deadline: 15 Dec. 2019
Registration Date: 3 Jan. 2020
Date of the Forum: 4-5 Jan. 2020

An introduction to SUSTech, research presentations, and tours inside SUSTech and Shenzhen City will be arranged during the Forum.

For more information on how to apply to join the forum, visit The application will close on 15 Dec. 2019. Successful applicants will receive the invitation before 20 Dec. 2019.

Call for Applications: 2020 Graduate Student Research Grants (15 Mar.)

The IEEE Computational Intelligence Society (CIS) funds scholarships to assist undergraduate, graduate and PhD students with the financial support to carry out their research during an academic break period. The primary intent of these scholarships is to cover the expenses related to a visit to another university, institute or research agency for collaboration with an identified researcher in the field of interest of the applicant. Funds can be used to cover travel expenses as well as certain living expenses (such as housing).The field of interest of applicants is open, but should be connected with identifiable component of the CIS (neural networks, fuzzy systems, or evolutionary computation). The deadline for the next round of applications is 15 Mar. 2020. More information on the scheme, along with the outputs of past grant holders, can be found on the CIS Graduate Student Research Grants webpage at

Call for Papers (Journal)

Call for Papers (Conference)

CIS Conferences

2020 IEEE Conference on Evolving and Adaptive Intelligent Systems (EAIS)
Bari, Italy
27-29 May 2020
2020 IEEE Conference on Games (CoG)
Higashiosaka, Japan
24-27 Aug. 2020
2021 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2021)
11-14 Jul. 2021
2022 IEEE World Congress on Computational Intelligence (IEEE WCCI 2022)
Padua, Italy
11-16 Jul. 2022
Leandro L. Minku
University of Birmingham, UK
{{my.Comm Preferences:default=edit me}}