Filtreler

2022 - 20243
Filtreleri Sıfırla

Ayarlar

search.filters.applied.f.publicationcategory: Konferans Öğesi - Uluslararası - Kurum Öğretim Elemanı × Konu: Adversarial machine learning ×

Arama Sonuçları

Listeleniyor 1 - 3 / 3
  • Küçük Resim
    Yayın
    Unreasonable effectiveness of last hidden layer activations for adversarial robustness
    (Institute of Electrical and Electronics Engineers Inc., 2022) Tuna, Ömer Faruk; Çatak, Ferhat Özgür; Eskil, Mustafa Taner
    In standard Deep Neural Network (DNN) based classifiers, the general convention is to omit the activation function in the last (output) layer and directly apply the softmax function on the logits to get the probability scores of each class. In this type of architectures, the loss value of the classifier against any output class is directly proportional to the difference between the final probability score and the label value of the associated class. Standard White-box adversarial evasion attacks, whether targeted or untargeted, mainly try to exploit the gradient of the model loss function to craft adversarial samples and fool the model. In this study, we show both mathematically and experimentally that using some widely known activation functions in the output layer of the model with high temperature values has the effect of zeroing out the gradients for both targeted and untargeted attack cases, preventing attackers from exploiting the model's loss function to craft adversarial samples. We've experimentally verified the efficacy of our approach on MNIST (Digit), CIFAR10 datasets. Detailed experiments confirmed that our approach substantially improves robustness against gradient-based targeted and untargeted attack threats. And, we showed that the increased non-linearity at the output layer has some ad-ditional benefits against some other attack methods like Deepfool attack.
  • Küçük Resim
    Yayın
    Segmentation based classification of retinal diseases in OCT images
    (Institute of Electrical and Electronics Engineers Inc., 2024) Eren, Öykü; Tek, Faik Boray; Turkan, Yasemin
    Volumetric optical coherence tomography (OCT) scans offer detailed visualization of the retinal layers, where any deformation can indicate potential abnormalities. This study introduced a method for classifying ocular diseases in OCT images through transfer learning. Applying transfer learning from natural images to Optical Coherence Tomography (OCT) scans present challenges, particularly when target domain examples are limited. Our approach aimed to enhance OCT-based retinal disease classification by leveraging transfer learning more effectively. We hypothesize that providing an explicit layer structure can improve classification accuracy. Using the OCTA-500 dataset, we explored various configurations by segmenting the retinal layers and integrating these segmentations with OCT scans. By combining horizontal and vertical cross-sectional middle slices and their blendings with segmentation outputs, we achieved a classification a ccuracy of 91.47% and an Area Under the Curve (AUC) of 0.96, significantly outperforming the classification of OCT slice images.
  • Küçük Resim
    Yayın
    Machine learning for adaptive modulation in medical body sensor networks using visible light communication
    (Institute of Electrical and Electronics Engineers Inc., 2024) Rizi, Reza Bayat; Forouzan, Amir Reza; Miramirkhani, Farshad; Sabahi, Mohamad Farzan
    In the context of medical body sensor networks that rely on visible light communication (VLC), adaptive modulation plays a crucial role. Despite VLC's advantages, challenges arise due to fluctuating signal strength caused by patient movement. To address this, we propose an adaptive modulation system that adjusts based on link conditions, specifically the signal-to-noise ratio (SNR). Our approach involves an uplink channel for feedback, allowing the receiver to select the appropriate modulation scheme based on measured SNR after noise mitigation. The analysis focuses on various medical situations and investigates machine learning algorithms. The study compares adaptive modulation based on supervised learning with that based on reinforcement learning. By implementing a bi-directional system with real-time modulation tracking, we demonstrate the effectiveness of adaptive VLC in handling environmental changes (interference and noise). Notably, the use of the Q-learning algorithm enables real-time adaptation without prior knowledge of the environment. Our simulation results show that photodetectors placed on the shoulder and wrist benefit significantly from this approach, experiencing improved performance.