Home > News&Eventes > Content

The Paper of an Undergraduate Student from EIC Accepted by ACM Multimedia 2021

Author:蔡普成、张嘉妮Time:2021-07-29Hits:

On July 4, 2021, the 29th ACM International Conference on Multimedia (ACM MM2021), a top international conference in the multimedia field, released the list of accepted papers in 2021. Towards Adversarial Patch Analysis and Certified Defense against Crowd Counting, the paper composed by Wu Qiming as the first author, an undergraduate student of class 2018 from EIC, was accepted by ACM Multimedia 2021.


The Poster of ACM MULTIMEDIA 2021


    The abstract

Crowd counting has drawn much attention due to its importance in safety-critical surveillance systems. Especially, deep neural network (DNN) methods have significantly reduced estimation errors for crowd counting missions. Recent studies have demonstrated that DNNs are vulnerable to adversarial attacks, i.e., normal images with human-imperceptible perturbations could mislead DNNs to make false predictions. In this work, we propose a robust attack strategy called Adversarial Patch Attack with Momentum (APAM) to systematically evaluate the robustness of crowd counting models, where the attacker's goal is to create an adversarial perturbation that severely degrades their performances, thus leading to public safety accidents (e.g., stampede accidents). Especially, the proposed attack leverages the extreme-density background information of input images to generate robust adversarial patches via a series of transformations (e.g., interpolation, rotation, etc.). We observe that by perturbing less than 6\% of image pixels, our attacks severely degrade the performance of crowd counting systems, both digitally and physically. To better enhance the adversarial robustness of crowd counting models, we propose the first regression model-based Randomized Ablation (RA), which is more sufficient than Adversarial Training (ADT) (Mean Absolute Error of RA is 5 lower than ADT on clean samples and 30 lower than ADT on adversarial examples). Extensive experiments on five crowd counting models demonstrate the effectiveness and generality of the proposed method.



About the author and his supervisor

In the context that undergraduate students are encouraged to join research groups and laboratories as early as possible, Zhou Pan, professor at the School of Cyber Science and Engineering at HUST and Wu Qiming’s supervisor, has provided a variety of opportunities for undergraduate students to carry out scientific research together with doctoral and graduate students. In this way, not only can undergraduate students’ research interest and research awareness be enhanced, but also the quality of talent training can be improved.

Wu Qiming, the first author of the paper and an undergraduate student of class 2018 from EIC at HUST, joined Professor Zhou Pan’s research group when he was only a freshman. When he encounters difficulties in research, Professor Zhou Pan is always willing to provide instructions and support for him. Moreover, Wang Binghui, assisting professor from Illinois Institute of Technology, and Doctor Li Ang from Duke University have also provided guidance during the process of paper composition, both of whom spoke highly of Wu Qiming. Wu Qiming is now in the research group of Wang Zhangyang, assisting professor from University of Texas at Austin, during summer vacation.


Wu Qiming


Introduction to ACM Multimedia

Since the founding of ACM SIGMM in 1993, ACM Multimedia has been the worldwide premier conference and a key world event to display scientific achievements and innovative industrial products in the multimedia field. For the first time in its history, ACM Multimedia 2021 will be held in Chengdu, the capital city of the Sichuan Province in China. At ACM Multimedia 2021, we will present an extensive program consisting of technical sessions covering all aspects of the multimedia field via oral, video and poster presentations, tutorials, panels, exhibits, demonstrations, workshops, doctoral symposium, multimedia grand challenge, brave new ideas on shaping the research landscape, open source software competition, and also an interactive arts program stimulating artists and computer scientists to meet and discover together the frontiers of artistic communication.


 

Address: 1037 Luoyu Road, Wuhan, China

Copyright©2017 Huazhong University of Science & Technology