International Conference on Image Analysis and Recognition

14th International Conference on
Image Analysis and Recognition

ICIAR 2017
July 5-7, 2017 – Montreal, Canada

Association for Image and Machine Intelligence

Invited Speakers

Terry Peters
Robarts Research Institute
Western University

The role of Ultrasound and Augmented reality for intra-cardiac beating heart interventions

Over the past half century, medical Imaging has grown in sophistication and its use has evolved well beyond diagnosis. Much effort been dedicated to minimizing invasiveness in surgical interventions, most of which has been achieved through developments in medical imaging, surgical navigation, visualization and display technologies. Image-guided procedures hold the promise to dramatically change the way therapies are delivered to many organs. This presentation provides an overview of developments in image-guided interventions with particular emphasis on cardiac interventions, and suggests strategies that will stimulate wider their adoption.

Download PPT [172 MB]

Bio »

Tien Bui
Department of Computer Science and Software Engineering
Concordia University

Matrix-Tensor Representation and Deep Learning in Large Scale Image Data Analysis

In this talk I will discuss two different perspectives in large scale image data analysis. The first is based on matrix/tensor representation framework where techniques such as sparse representation, low rank approximation, dictionary learning, robust principal component analysis, and multi-linear principal component analysis have received much attention as powerful tools for image/video processing, pattern recognition, and computer vision. The second is based on deep learning techniques such as convolution nets, recurrent nets, and the restricted Boltzmann machines. The talk is an attempt to make connections between the two. Some of the latest works in large scale image data processing will be presented.

Download PDF [14 MB]

Bio »

Andrew K.C. Wong
Systems Design Engineering
University of Waterloo

Discovering Deep Knowledge from Biosequence and Temporal Sequence Data

This talk presents a novel method P2K (Pattern-to- Knowledge) with intriguing findings that deep knowledge could be discovered from biological and temporal sequence data without reliance on explicit prior knowledge. From biosequence data, it discovers the “what” and “where” of crucial functional units, elements or mechanisms related explicitly to the physio-chemical environment. Its outcomes can be validated by the work of domain experts reported in literature or through new tests/experiments. It enhances the experts’ insights and efficiency by shortening the search process and improving predictive analysis. It will open an avenue to discover deep knowledge from big sequence data for biology, drug discovery and medical research. P2K has also been applied to temporal sequence patterns with varying magnitude and time delays to discover a wide range of local relations along/across mixed-mode time series without relying explicitly on prior knowledge or models. Incorporated in P2K is a new method for predicting rare events in imbalanced data that has great potential for abnormal financial trend forecast and fault detection. It meets a new challenge in the era of big data.

Bio »


ICIAR - International Conference on Image Analysis and Recognition
Send your comments to