Improving Perceptual Quality of Adversarial Images Using Perceptual Distance Minimization and Normalized Variance Weighting

2022-02-28
Karlı, Berat Tuna
Şen, Deniz
Temizel, Alptekin
36th AAAI Conference on Artificial Intelligence, Adversarial Machine Learning and Beyond Workshop

Suggestions

Improving Perceptual Quality of Spatially Transformed Adversarial Examples
Aydın, Ayberk; Temizel, Alptekin; Department of Modeling and Simulation (2022-8)
Deep neural networks are known to be vulnerable to additive adversarial perturbations. The amount of these additive perturbations are generally quantified using Lp metrics over the difference between adversarial and benign examples. However, even when the measured perturbations are small, they tend to be noticeable by human observers since Lp distance metrics are not representative of human perception. Spatially transformed examples work by distorting pixel locations instead of applying an additive perturba...
Improving data freshness in random access channels
Atabay, Doğa Can; Uysal, Elif; Department of Electrical and Electronics Engineering (2019)
The conventional network performance metrics such as throughput and delay do not accurately reflect the needs of some applications. Age of information (AoI) is a newly proposed metric that indicates the freshness of information from the receiver’s perspective. In this work, a network of multiple transmitter devices continuously updating a central station over an error-free multiaccess channel is studied. The average AoI expressions are derived for Round-Robin, Slotted ALOHA, and a proposed random access str...
Improving classification performance of endoscopic images with generative data augmentation
Çağlar, Ümit Mert; Temizel, Alptekin; Department of Modeling and Simulation (2022-2-8)
The performance of a supervised deep learning model is highly dependent on the quality and variety of the images in the training dataset. In some applications, it may be impossible to obtain more images. Data augmentation methods have been proven to be successful in increasing the performance of deep learning models with limited data. Recent improvements on Generative Adversarial Networks (GAN) algorithms and structures resulted in improved image quality and diversity and made GAN training possible with lim...
Improving operational performance of antennas on complex platforms by arranging their placements
Bayseferoğulları, Can; Dural Ünver, Mevlüde Gülbin; Department of Electrical and Electronics Engineering (2010)
The aim of this thesis is to improve the operational performance of the communication antennas mounted on complex platforms such as aircrafts and warships by arranging placements of these antennas. Towards this aim, primarily, in order to gain insight on the influence of geometrically simple structures composing the platform on antenna performance, a quarter wavelength monopole antenna placed at the center of a finite square ground plane is studied by using uniform Geometrical Theory of Diffraction (GTD). B...
Improving pattern quality in web usage mining by using semantic information
Karagöz, Pınar (Springer Science and Business Media LLC, 2012-03-01)
Frequent Web navigation patterns generated by using Web usage mining techniques provide valuable information for several applications such as Web site restructuring and recommendation. In conventional Web usage mining, semantic information of the Web page content does not take part in the pattern generation process. In this work, we investigate the effect of semantic information on the patterns generated for Web usage mining in the form of frequent sequences. To this aim, we developed a technique and a fram...
Citation Formats
B. T. Karlı, D. Şen, and A. Temizel, “Improving Perceptual Quality of Adversarial Images Using Perceptual Distance Minimization and Normalized Variance Weighting,” presented at the 36th AAAI Conference on Artificial Intelligence, Adversarial Machine Learning and Beyond Workshop, Vancouver, Kanada, 2022, Accessed: 00, 2022. [Online]. Available: https://hdl.handle.net/11511/97506.