Do you have an account?

Judgment panel

Important Dates

Submission of papersFebruary 11, 2018

Proposals for tutorials and workshopsJanuary 20, 2018

Notification of paper and proposals acceptanceFebruary 24, 2018

Submission of camera-readyMarch 10, 2018

Date of conferenceApril 08, 2018

Viewers' statistics

Online 14

Today 140

Totall 186577

Workshops

Workshops

Dr. Alireza Behrad

Associate Professor, Shahed University

WorkshopTitle: Forgery detection and Forensics in digital images and videos

 Abstract: In recent years the availability of inexpensive photo-editing software has made the manipulation, editing and copying of the images and videos easier. Consequently, the validation of image or video content has become more difficult. This implies that image and video contents cannot be used as legal evidence. To handle this problem, considerable research efforts have been devoted to determine the originality of image, video or media contents. The aim of digital image and video forensics is to determine whether an image or video content is original or modified. Several algorithms have been proposed for image and video forensics and forgery detection.

Image forensics based on image acquisition artifacts is one of the basic approaches for image forensics. The most commonly used algorithm of this category is based on the noise of acquisition pipeline which is called photo response non-uniformity (PRNU) noise. It is shown that the PRNU for each camera is unique and can be employed for image forensics. The main problem of PRNU based algorithms is the accurate estimation of PRNU which needs several images of the same scene. The calculation of the PRNU for compressed images may not be enough accurate.

Image compression and coding is inevitable part of image and video formation process for most current image cameras and image editing software. Because of the lossy nature of compression process, the process has its own impact on the resultant image or video. Some image and video forensic algorithms utilized coding or compressing fingerprint to detect image forgery. Variety of algorithms of this group is focused on double compression detection. Statistical properties of quantized DCT coefficients using Benford’s law is one the most commonly used approach for double compression detection.  

Image formation is the process of projecting 3D world points into 2D image pixels using perspective geometry model. Additionally, the illumination of the scene is converted into image intensity through a lens system. Human brain is not generally good at estimating illumination physic and geometric constraints in the image. As a consequence, the forged image or video generally includes some geometric or physical inconsistency. Algorithms based on geometric and physic inconsistencies use illumination or geometric inconsistency in the image or video for forgery detection.

Any process which is applied to image for the purpose of image processing, image enhancement or image forgery is called image editing operator. Some editing operators are used to enhance the quality of the original image and called “innocent”. However some other which are called “malicious” are used to change the semantic content of the image and categorized as image forgery algorithms. The most important malicious editing algorithms are copy-move and cut-and-paste attacks. Algorithms based on editing fingerprints use scene or signal level inconsistency generated by editing operators to detect image forgery.

In this workshop scientific basics and some recent algorithms for forgery detection and image and video forensics are discussed.

Registration Deadline: 17 March 2018

How to register? Send profile to Email: symp@qiau.ac.ir


Dr. Khalil Alipour

Assistant Professor, Tehran University

WorkshopTitle: Wheeled Mobile Robots: Modeling and Control

 Abstract: In contrast to fixed-base robot arms, mobile robots have infinite workspace. Such robotic systems contain various types. Among various mobile robots, Wheeled Mobile Robots (WMRs) are simple, have high speeds and are highly energy-efficient. In this workshop, first, various kinds of mobile robots are introduced and their advantages/disadvantages are discussed. Then, various kinds of wheels and WMR structures are examined. Since simulation, optimization and control of the WMRs require their model, kinematics and dynamics of them are derived. Next, several structures of their dynamics model are. The mobility of WMRs is discussed on the basis of the pure rolling assumption for the wheels. The model of the WMRs is extracted for both cases in which the wheels do not slip and once there exist slipping and skidding between the wheels and the ground. Next, various control strategies to adjust the motion of WMRs are explained. Finally, some future research directions are introduced.

 

Registration Deadline: 17 March 2018

How to register? Send profile to Email: symp@qiau.ac.ir

Count down

0

Day

Poster of conference

Sponsor