Recently, adversarial machine learning attacks have posed serious security threats against practical audio signal classification systems, including speech recognition, speaker recognition, music copyright detection. Most existing studies have mainly focused on ensuring the effectiveness of attacking an audio signal classifier via creating a noise-like perturbation on the original signal, which remains a gap in preserving the human perception of adversarial audios. This paper presents a novel perspective to create adversarial audios by integrating the human perception model into the attack formulation to generate well-perceived adversarial examples. Different from conventional approaches which primarily focused on using$L_{p}$norm to preserve the audio quality, we adopt a human study to understand how human participants react to different types of music perturbations, build a Siamese Neural Network (SNN) based model to characterize the human perception. The new findings of the human perception study guide us to formulate a new computationally efficient, multiple-feature-based perception-aware (CEMF-PA) attack, which manipulates different audio signal features to find an optimal perturbed music signal against music copyright detection. This novel attack vector opens a new door to generating highly effective, well-perceived adversarial audio signals via manipulating the auditory features. Experimental results show that the proposed attack is effective against YouTube’s copyright detection. Finally, we propose the defense strategy design to make the copyright detection more robust to adversarial music signals generated by the CEMF-PA attack.

Three-dimensional (3D) cloud retrievals are critical for understanding their impact on climate and other applications such as aviation safety, weather prediction, and remote sensing. However, obtaining high-resolution and accurate vertical representation of clouds remains unsolved due to the limitations imposed by satellite instrumentation, viewing conditions, and the complexity of cloud dynamics. Cloud masks are essential for comprehending various cloud vertical properties, but deriving accurate 3D cloud masks from 2D satellite imagery data is a challenging task. To tackle these challenges, we introduce a physics-informed loss function for training deep learning models that can extend 2D cloud images into 3D cloud masks. The proposed loss, calledCloudMask Loss, is composed of two domain knowledge-informed loss terms: one for evaluating cloud position and thickness, and the other for measuring the number of layers. By combining these loss terms, we improve the trainability of the deep learning models for more accurate and meaningful results. We apply the proposed loss function to different neural networks and demonstrate significant improvements in multi-layer cloud mask reconstruction. Utilizing the same neural network architecture, our proposed loss outperforms standard binary cross-entropy loss in terms of multi-layer cloud classification accuracy, number of layers accuracy, and thickness mean absolute error (MAE). The proposed loss function can be readily integrated into various neural network architectures, resulting in substantial performance gains in 3D cloud mask generation.

Clouds at different altitudes play different roles in Earth’s climate. Comprehensive understanding of overlapping clouds is important for climate and weather prediction. The East Pacific region is where El Niño and La Niña originate and where multi-layer clouds frequently occur. The overlap of clouds at different altitudes in this region increases the classification complexity for cloud-based climatological studies. Unlike prior work in cloud layer classification that assumes single layer or two-layer of clouds, in this work, we consider multi-layer cloud classification with 8 cloud-level classes (clear-sky, high, middle, low, high+middle, high+low, middle+low, high+middle+low). We develop and analyze machine learning models on features extracted from satellite images from the East Pacific regions collected by GOES Advanced Baseline Imager (ABI). These are used to classify CloudSat/CALIPSO observed multi-layer clouds. Due to the imbalanced nature of the data, we investigate the adoption of conventional resampling methods, as well as deep learning methods with data augmentation. In our experiments, we utilize the random forest classifier and Multilayer perceptron classifier with data augmentation methods to reduce the class imbalance during training. With these approaches, we achieve a classification accuracy of 83.6% without exploiting any ancillary information.

Previous adversarial audio attacks have mainly focused on ensuring the effectiveness of attacking an audio signal classifier via creating a small noise-like perturbation on the original signal. It is still unclear if an attacker is able to create audio signal perturbations that can be well perceived by human beings in addition to its attack effectiveness. In this work, we formulate the adversarial attack against music signals as a new perception-aware attack framework, which integrates human study into adversarial attack design. Specifically, we invite human participants to rate their perceived deviation based on pairs of original and perturbed music signals, and reverse-engineer the human perception process by regression analysis to predict the human-perceived deviation given a perturbed signal. The perception-aware attack is then formulated as an optimization problem that finds an optimal perturbation signal to minimize the prediction of perceived deviation from the regressed human perception model. Experiments show that the attack produces adversarial music with significantly better perceptual quality than prior work against YouTube's copyright detector.

DeepFake represents a category of face-swapping attacks that leverage machine learning models such as autoen-coders or generative adversarial networks. Although the concept of the face-swapping is not new, its recent technical advances make fake content (e.g., images, videos) imperceptible to Humans. Various detection techniques for DeepFake attacks have been explored. These methods, however, are passive measures against DeepFakes as they are mitigation strategies after the high-quality fake content is generated. This work aims to take an offensive measure to impede the generation of high-quality fake images or videos. Specifically, we propose to use novel transformation-aware adversarially perturbed faces as a defense against GAN-based DeepFake attacks, which leverages differentiable random image transformations during the generation. We also propose an ensemble-based approach to enhance the defense robustness against GAN-based DeepFake variants under the black-box setting. We show that training a DeepFake model with adversarial faces can lead to a significant degradation in the quality of synthesized faces.

Smart home devices have brought us many benefits such as advanced security, convenience, and entertainment. However, these devices also have made unintended consequences like giving ultimate power for devices' owners over their intimate partners in the same household which might lead to tech-facilitated domestic abuse (tech-abuse) as recent research has shown. In this paper, we systematize findings on tech-abuse in smart homes. We show that domestic abuse and Intimate Partner Violence (IPV) in smart homes is more effective and less risky for abusers. Victims find it more harmful and more challenging to protect themselves from. We articulate a comprehensive analysis of all the phases of abuse in smart homes and categorize risks and needs in each phase by designing a unified analytical framework. Technical analysis of current smart home technologies is conducted to shed light upon their limitations. We also summarize recent recommendations to combat tech-abuse in smart homes and focus on their potentials and shortcomings. Unsurprisingly, we find that many recommendations conflict with each other due to a lack of understanding of phases of abuse in smart homes. Desirable properties to design abuse-resistant smart home devices are proposed for all the phases of abuse. The research community benefits from our analysis and recommendations to move forward with a focus on filling the blind spots of existing smart home devices' safety measures and building appropriate safety measures that consider tech-abuse threats in smart homes.

We consider a secondary multi-hop cognitive radio network with decentralized control that operates cognitively to coexist with primary users. We propose a new spread-spectrum management paradigm, in which, unlike mainstream dynamic spectrum access research, digital waveforms are designed to occupy the entire available spectrum, and to adaptively track the interference profile at the receiver to maximize the link capacity while avoiding interference to primary users. In this context, we study the problem of maximizing the network throughput of a multi-hop network through joint routing and spread-spectrum channelization. We first propose a centralized formulation of the network control problem. We then propose an algorithm that can be seen as a distributed localized approximation of the throughput-maximizing policy. We refer to the proposed jointly-designed routing and code-division channelization algorithm as ROCH (Routing and cOde-division CHannelization). Specifically, power and spreading code are jointly selected to maximize the pre-detection secondary \mathrm{SINR} while providing quality of service guarantees to on-going primary and secondary transmissions, while the routing algorithm dynamically selects relays based on the network traffic dynamics and on the achievable data rates on different secondary links. We study the throughput and delay performance of ROCH through a extensive simulation experiments, which demonstrate the appeal of the proposed framework through significant performance gains compared to baseline solutions.

The problem of optimal resource allocation to share high-quality multimedia content in cognitive ad hoc networks with cooperative relays is addressed in this paper. Cooperative transmission is a promising technique to increase the capacity of wireless links by exploiting spatial diversity without multiple antennas at each node. However, mainstream research in this field focuses on optimizing physical layer performance measures, with little consideration for application-specific and network-wide performance measures. In this paper, the problem of joint video encoding rate control, power control, relay selection and channel assignment is formulated as a mixed-integer nonlinear problem(MINLP), and a solution algorithm based on a combination of the branch and bound framework and convex relaxation techniques is then proposed. The proposed solution jointly allocates channel, power, video encoding rate, and relay nodes for secondary users to maximize the video quality under the constraints posed by delay-sensitive video applications. Performance evaluation results show that cognitive networks with cooperative relaying can provide considerably higher video quality (in terms of the average peak signal-to-noise ratio (PSNR)) than solutions that do not rely on cooperation or without dynamic spectrum allocation.

Throughput maximization is one of the main challenges in cognitive radio ad hoc networks, where the availability of local spectrum resources may change from time to time and hop-by-hop. Cooperative transmission exploits spatial diversity without multiple antennas at each node to increase capacity with reliability guarantees. This idea is particularly attractive in wireless environments due to the diverse channel quality and the limited energy and bandwidth resources. With cooperation, source node and relay node cooperatively transmit data to the destination. In such a virtual multiple antenna transmission system, the capacity of the cooperative link is much larger than that of the direct link from source to destination. In this paper, we will study decentralized and localized algorithms for joint dynamic routing, relay assignment, and spectrum allocation under a distributed and dynamic environment.

A cognitive radio network with decentralized control (i.e., a cognitive ad hoc network) is considered in this demonstration. The demo implements a decentralized and localized algorithm for through put maximization through joint routing and interference-avoiding waveform selection. The algorithm adapts to time-varying traffic demands, interference profile, and network topology to locally maximize the achievable data rate while avoiding harmful interference to co-located primary or secondary users. The prototype is based on a cross-layer protocol stack implemented in Python, which leverages GNU Radio for adaptive signal generation on a USRP2 software-defined-radio platform.

Throughput maximization is one of the main challenges in cognitive radio ad hoc networks, where local spectrum resources may change from time to time and hop-by-hop. For this reason, a cross-layer opportunistic spectrum access and dynamic routing algorithm for cognitive radio networks is proposed, called ROSA (ROuting and Spectrum Allocation algorithm). Through local control actions, ROSA aims at maximizing the network throughput by performing joint routing, dynamic spectrum allocation, scheduling, and transmit power control. Specifically, the algorithm dynamically allocates spectrum resources to maximize the capacity of links without generating harmful interference to other users while guaranteeing bounded BER for the receiver. In addition, the algorithm aims at maximizing the weighted sum of differential backlogs to stabilize the system by giving priority to higher-capacity links with high differential backlog. The proposed algorithm is distributed, computationally efficient, and with bounded BER guarantees. ROSA is shown through discrete-event packet-level simulations to outperform baseline solutions leading to a high throughput, low delay, and fair bandwidth allocation.