Keynote Speakers/主讲嘉宾

Keynote Speakers/主讲嘉宾

Keynote Speeches

Assist. Prof. Peng LU 

Hong Kong Polytechnic University

Title: High-precision control for ground and flying robotics

Abstract: The control of aircraft has become more challenging due to the fact that the aircraft structure is becoming increasingly complex and flexible materials are being used. The need of developing more advanced motion planning and control techniques is increasing. State-of-the-art motion planning and control techniques are challenged by model uncertainties as well as disturbances. 

For ground robotics, research on control of the whole-body dynamics has aroused a lot of attention but also met with significant challenges. Take hydraulic robots for example, the dynamics of these robots are fast and the control of these robots is affected by the friction in the hydraulic actuators. Therefore, more advanced control techniques should be designed to improve their performance and also make them more autonomous.

A. Prof. Sathish Kumar Selvaperumal

Asia Pacific University of Technology and Innovation, Malaysia

Title: Emerging Technologies – Deep Learning

Abstract: Recent research trends and technology based on Gartner’s Hype cycle (2014 to 2018) is exposed as the Journey to Digital Business. Introduction to few mega trends such as IoT, Deep Learning and 5G is exposed. A research outcome about Car make model and recognition using computer vision that could distinguish between different models of cars including manufacturing year of the same make and identify them from the image. A total of three deep learning neural network systems were used in this research. Deep learning offers outstanding solution with high accuracy and many companies are investing in this artificial intelligence research.

Oral Presentation

Xiaocheng Chang (常小铖)

Wuhan University, China

Title: Comprehensive Analysis of Conductor Selection

With the continuous improvement of people's living standard, the consumption of electricity also increased, which requires a larger capacity of the conductor for electricity transmission. Therefore, comprehensive selection of conductor type made higher requirements. Based on the analysis, this paper divided the basic indication of conductor selection into two types: cost type and efficiency type, then each category divided into a number of specific indicators based on needs, finally, using the grey relational analysis method to compare and analyzes the cost indication and efficiency indication of conductors, drawing the optimization scheme of conductor selection, and using it to guide the type selection of conductor type.

Yali Ren(任雅丽)

Southeast University, China

Title: A GP-ELM Gesture Recognition Algorithm for Tactile Animation

Artificial Neural Network (ANN) is one of the most important techniques in gesture recognition with high accuracy, but its slow response limited its real-time applications. Extreme Learning Machine (ELM), a novel ANN with fast learning ability, has the potential to solve the above problem. The extant recognition methods based on ELM focus on Local Variant Features (LVF) of gesture images and a small number of data are used to represent many features. However, the locality of these features may lead to an inadequate representation of real world information. In this paper, a new algorithm based on Global Pixels (GP) and ELM is proposed. It includes two parts: image pre-processing and GP-ELM identification. Through the processing of segmentation, filtering and standardization, a raw gesture image is transformed into a binarization and unified-size image. And then the GP of the picture are used as the ELM inputs to be trained or identified. The algorithm using GP instead of LVF for training simplifies the pre-processing without losing information. Experiment results show that the average recognition accuracy can reach 96.67%, while the recognition time is only 0.041 milliseconds. With more hidden nodes, the accuracy can be improved to be 100% with negligible increase in recognition time.