No abstract

As the popularity of voice user interface (VUI) exploded in recent years, speaker recognition system has emerged as an important medium of identifying a speaker in many security-required applications and services. In this paper, we propose the first real-time, universal, and robust adversarial attack against the state-of-the-art deep neural network (DNN) based speaker recognition system. Through adding an audio-agnostic universal perturbation on arbitrary enrolled speaker's voice input, the DNN-based speaker recognition system would identify the speaker as any target (i.e., adversary-desired) speaker label. In addition, we improve the robustness of our attack by modeling the sound distortions caused by the physical over-the-air propagation through estimating room impulse response (RIR). Experiment using a public dataset of 109 English speakers demonstrates the effectiveness and robustness of our proposed attack with a high attack success rate of over 90%. The attack launching time also achieves a 100× speedup over contemporary non-universal attacks.

Chemotherapy is a strategy for patients with advanced prostate cancer, especially those with castration-resistant prostate cancer. Prostate cancer stem cells (PCSCs) are believed to be the origin of cancer recurrence following therapy intervention, including chemotherapy. The mechanisms underlying the chemoresistance of PCSCs are still poorly understood. In the present study, fluorescence-activated cell sorting was used to isolate PCSCs from LNCaP and PC3 cell lines. 3-(4,5-Dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide was used to measure the cell viability. Quantitative real-time PCR and western blotting were utilized to evaluate the mRNA and protein levels. ShRNA was employed to knock down target gene expression. Chromatin immunoprecipitation (ChIP) was performed to explore the detailed mechanism underlying ABCC1 expression. Our results revealed that the sorted PCSCs showed enhanced chemoresistance ability than matched non-PCSCs. Protein level of activated form of NOTCH1(ICN1) was significantly higher in PCSCs. Inhibition of NOTCH1 with shRNA could decrease ABCC1 expression, and improve chemosensitivity in PCSCs. Finally, ChIP-PCR showed ICN1 could directly bind to the promoter region of ABCC1. In conclusion, NOTCH1 signaling could transactivate ABCC1, resulting in higher chemoresistance ability of PCSCs, which might be one of the important mechanisms underlying the chemoresistance of PCSCs.

The study of adversarial vulnerabilities of deep neural networks (DNNs) has progressed rapidly. Existing attacks require either internal access (to the architecture, parameters, or training set of the victim model) or external access (to query the model). However, both the access may be infeasible or expensive in many scenarios. We investigate no-box adversarial examples, where the attacker can neither access the model information or the training set nor query the model. Instead, the attacker can only gather a small number of examples from the same problem domain as that of the victim model. Such a stronger threat model greatly expands the applicability of adversarial attacks. We propose three mechanisms for training with a very small dataset (on the order of tens of examples) and find that prototypical reconstruction is the most effective. Our experiments show that adversarial examples crafted on prototypical auto-encoding models transfer well to a variety of image classification and face verification models. On a commercial celebrity recognition system held by clarifai.com, our approach significantly diminishes the average prediction accuracy of the system to only 15.40%, which is on par with the attack that transfers adversarial examples from a pre-trained Arcface model. Our code is publicly