- 著者
- Hyun KWON Changhyun CHO Jun LEE
- 出版者
- The Institute of Electronics, Information and Communication Engineers
- 雑誌
- IEICE TRANSACTIONS on Information and Systems (ISSN:09168532)
- 巻号頁・発行日
- vol.E105-D, no.11, pp.1880-1889, 2022-11-01
- 被引用文献数
- 2
Deep neural networks (DNNs) provide excellent services in machine learning tasks such as image recognition, speech recognition, pattern recognition, and intrusion detection. However, an adversarial example created by adding a little noise to the original data can result in misclassification by the DNN and the human eye cannot tell the difference from the original data. For example, if an attacker creates a modified right-turn traffic sign that is incorrectly categorized by a DNN, an autonomous vehicle with the DNN will incorrectly classify the modified right-turn traffic sign as a U-Turn sign, while a human will correctly classify that changed sign as right turn sign. Such an adversarial example is a serious threat to a DNN. Recently, an adversarial example with multiple targets was introduced that causes misclassification by multiple models within each target class using a single modified image. However, it has the weakness that as the number of target models increases, the overall attack success rate decreases. Therefore, if there are multiple models that the attacker wishes to attack, the attacker must control the attack success rate for each model by considering the attack priority for each model. In this paper, we propose a priority adversarial example that considers the attack priority for each model in cases targeting multiple models. The proposed method controls the attack success rate for each model by adjusting the weight of the attack function in the generation process while maintaining minimal distortion. We used MNIST and CIFAR10 as data sets and Tensorflow as machine learning library. Experimental results show that the proposed method can control the attack success rate for each model by considering each model's attack priority while maintaining minimal distortion (average 3.95 and 2.45 with MNIST for targeted and untargeted attacks, respectively, and average 51.95 and 44.45 with CIFAR10 for targeted and untargeted attacks, respectively).
- 著者
- Yong Hyun Kwon Jung Won Kwon Sang Young Park Mi Young Lee Sung Ho Jang Chung Sun Kim
- 出版者
- The Society of Physical Therapy Science
- 雑誌
- Journal of Physical Therapy Science (ISSN:09155287)
- 巻号頁・発行日
- vol.23, no.4, pp.635-638, 2011 (Released:2011-09-22)
- 参考文献数
- 23
- 被引用文献数
- 1
[Purpose] Pusher syndrome, which is a disorder of postural balance that occurs in hemiparetic stroke patients, is characterized by a particular tendency to strongly push toward the hemiparetic side. The purpose of this study was to investigate whether stroke patients with pusher syndrome have ideomotor apraxic behavior. [Subjects] Fifteen stroke patients with pusher syndrome and 31 stroke patients without pusher syndrome were recruited. [Methods] All subjects were tested with two tests assessing ideomotor apraxia of movements of the upper and lower limbs. Each test included 12 items of movements, which required the subjects to reproduce movements by imitation after presentation. [Results] Patients with pusher syndrome had significantly lower ideomotor apraxia scores in all of the upper and lower limbs than patients without pusher syndrome. A significant difference was observed between the two groups in the existence of neglect. [Conclusion] We found that patients with pusher syndrome had more severe apraxic disorder in all of the upper and lower limbs than patients without pusher syndrome. Pusher syndrome may be attributable to disabilities in motor planning and execution, which are required to compensate for the partial damage to the postural control system.