In real-world environments, human speech is usually distorted by both reverberation and background noise, which have negative effects on intelligibility quality. They also cause performance degradation in many technology applications, such as automatic recognition. Therefore, the dereverberation denoising problems must be dealt with daily listening environments. this paper, we propose to perform using supervised learning, approach then extended address denoising. Deep neural networks are...

Vehicles face growing security threats as they become increasingly connected with the external world. Hackers, researchers, and car hobbyists have compromised keys used by electronic control units (ECUs) on vehicles, modified ECU software, hacked wireless transmissions from vehicle key fobs tire monitoring sensors, using low-cost commercially available tools. However, most damaging to vehicles are only emerging. One such threat is malware, which can infect in a variety of ways cause severe...

We have experimentally demonstrated a decoy-state quantum key distribution scheme (QKD) with heralded single-photon source based on parametric down-conversion. used one-way Bennett-Brassard 1984 protocol four states and one-detector phase-coding scheme, which is immune to recently proposed time-shift attacks, photon-number splitting can also be proven secure against Trojan horse attacks any other standard individual or coherent attacks. In principle, the setup tolerate highest losses it give...

Bug fixing is an essential activity in the software maintenance, because most of systems have unavoidable defects. When new bugs are submitted, triagers to find and assign appropriate developers fix bugs. However, if at first assigned inappropriate developers, they may later be reassigned other developers. That increases time cost for Therefore, finding becomes a key bug resolution. report, it necessary decide how quickly report should addressed. Thus, severity important factor fixing. In...

Recent years have witnessed the sharp increase of malicious apps that steal users' personal information. To address concerns about privacy risks, more and are accompanied with policies written in natural language because it is difficult for users to infer an app's behaviors according required permissions. However, little known whether these trustworthy or not. It worth noting a questionable policy may result from careless preparation by app developer intentional deception attacker. In this...

In daily listening environments, human speech is often degraded by room reverberation, especially under highly reverberant conditions. Such degradation poses a challenge for many processing systems, where the performance becomes much worse than in anechoic environments. To combat effect of we propose monaural (single-channel) dereverberation algorithm using temporal convolutional networks with self attention. Specifically, proposed system includes self-attention module to produce dynamic...

Human listeners often have difficulties understanding speech in the presence of background noise real world. Recently, supervised learning based enhancement approaches achieved substantial success, and show significant improvements over conventional approaches. However, existing try to minimize mean squared error between enhanced output pre-defined training target (e.g., log power spectrum clean speech), even though purpose such is improve noise. In this paper, we propose a new deep neural...

In the real world, speech is usually distorted by both reverberation and background noise. such conditions, intelligibility degraded substantially, especially for hearing-impaired (HI) listeners. As a consequence, it essential to enhance in noisy reverberant environment. Recently, deep neural networks have been introduced learn spectral mapping corrupted speech, shown significant improvements objective metrics automatic recognition score. However, listening tests not yet any benefit. this...

As defects in software modules may cause product failure and financial loss, it is critical to utilize defect prediction methods effectively identify the potentially defective for a thorough inspection, especially early stage of development lifecycle. For an upcoming version project, practical employ historical labeled data prior versions within same project conduct on current version, i.e., Cross-Version Defect Prediction (CVDP). However, dynamic evolution process that distribution (such as...

Learning-based program repair has achieved good results in a recent series of papers. Yet, we observe that the related work fails to some bugs because lack knowledge about 1) application domain being repaired, and 2) fault type repaired. In this paper, solve both problems by changing learning paradigm from supervised training self-supervised an approach called SelfAPR. First, SelfAPR generates samples on disk perturbing previous version enforcing neural model capture project-specific...

Person-job fit is the core technique of online recruitment platforms, which can improve efficiency by accurately matching job positions with seekers. Existing works mainly focus on modeling unidirectional process or overall matching. However, a two-way selection process, means that both candidate and employer involved in interaction should meet expectation each other, instead unilateral satisfaction. In this paper, we propose dual-perspective graph representation learning approach to model...

Smartphones have been widely used in recent years due to their capabilities of communication and multimedia processing, thus they also become attack targets malware. Threat malicious software has an important factor the safety smartphones. Android is most popular open-source smartphone operating system its permission declaration access control mechanisms can’t detect behavior In this paper, a new signature based malware detection framework named RobotDroid using SVM active learning algorithm...

Developers increasingly rely on API tutorials to facilitate software development. However, it remains a challenging task for them discover relevant tutorial fragments explaining unfamiliar APIs. Existing supervised approaches suffer from the heavy burden of manually preparing corpus-specific annotated data and features. In this study, we propose novel unsupervised approach, namely Fragment Recommender APIs with PageRank Topic model (FRAPT). FRAPT can well address two main challenges lying in...

With the fast penetration of commercial Virtual Reality (VR) and Augmented (AR) systems into our daily life, security issues those devices have attracted significant interests from both academia industry. Modern VR/AR typically use head-mounted (i.e., headsets) to interact with users, often store private user data, e.g., social network accounts, online transactions or even payment information. This poses threats, since in practice headset can be potentially obtained accessed by...

A privacy policy is a statement informing users how their information will be collected, used, and disclosed. Failing to provide correct may result in fine. However, writing tedious error-prone, because the author not well understand source code, which could written by others (e.g., outsourcing), or does know internals of third-party libraries without codes. In this paper, we propose develop novel system named AutoPPG automatically construct readable descriptions facilitate generation for...

A privacy policy is a statement informing users how their information will be collected, used, and disclosed. Failing to provide correct may result in fine. However, writing tedious error-prone, because the author not understand source code well as it could have been written by others (e.g., outsourcing), or does know internal working of third-party libraries used. In this paper, we propose develop novel system named AutoPPG automatically construct readable descriptions facilitate generation...