With the wide use of smartphones, more private data are collected and saved in smartphones. This raises higher requirements for secure effective user authentication scheme. Continuous leverages behavioral biometrics as identity information shows promising characteristics verification a continuous passive means. However, most studies require users to operate smartphones specific mobile application or perform user-defined touch operations. paper on wild, where it is hard characterize touching...

This paper aims to predict the apps a user will open on her mobile device next. Such an information is essential for many smartphone operations, e.g., app pre-loading and content pre-caching, save energy. However, it hard build explicit model that accurately depicts affecting factors their mechanism of time-varying usage behavior. presents deep reinforcement learning framework, named as DeepAPP, which learns model-free predictive neural network from historical data. Meanwhile, online...

Touch behavior biometric has been widely studied for continuous authentication on mobile devices, which provides a more secure in an implicit process. However, the existing touch biometric-based systems suffer from two issues. First, representation methods are hard to characterize operations under complex usage context. Second, accuracy of models is inclined degrade over time long-term real-life scenario due change data distribution caused by varying behavior. Toward this end, article, we...

Map matching for cellular data is to transform a sequence of cell tower locations trajectory on road map. It an essential processing step many applications, such as traffic optimization and human mobility analysis. However, most current map approaches are based Hidden Markov Models (HMMs) that have heavy computation overhead consider high-order information. This paper presents fast framework data, named DMM, which adopts recurrent neural network (RNN) identify the most-likely roads given...

This paper aims to predict a set of apps user will open on her mobile device in the next time slot. Such an information is essential for many smartphone operations, e.g., app pre-loading and content pre-caching, improve experience. However, it hard build explicit model that accurately captures complex environment context predicts at one time. presents deep reinforcement learning framework, named as DeepAPP, which learns model-free predictive neural network from historical usage data....

Continuous authentication, which provides identity verification using behavioral biometrics in an implicit and transparent manner, has shown potentials for protecting privacy. As the most common way of human-computer interaction, touch behavior pattern each user been proven distinctive widely adopted continuous authentication. However, based solutions rely on touchscreen signals obtained from high-level application programming interfaces, are hard to characterize fine-grained appearance...

This paper presents a novel map matching framework that adopts deep learning techniques to sequence of cell tower locations trajectory on road network. Map is an essential pre-processing step for many applications, such as traffic optimization and human mobility analysis. However, most recent approaches are based hidden Markov models (HMMs) or neural networks hard consider high-order location information heuristics observed from real driving scenarios. In this paper, we develop reinforcement...

Retrieving similar trajectories from a large trajectory dataset is important for variety of applications, like transportation planning and mobility analysis. Unlike previous works based on fine-grained GPS trajectories, this paper investigates the feasibility identifying cellular data observed by mobile infrastructure, which provide more comprehensive coverage. To handle localization errors low sample rates data, we develop holistic system, cellSim, seamlessly integrates map matching search....

Smartphones have become the most important devices for users to communicate and interact with different forms of media, at same time stored a large amount sensitive private data. The security protection such data has increasingly critical. As sensor technology rapid developed, diversity sensors on smartphones greatly increased (e.g., motion touchscreen sensor), empowering provide continuous implicit user authentication by capturing behavioral biometrics. Unfortunately, it remains challenge...

In a refrigeration unit, the amount of refrigerant has substantial influence on entire system. To predict in refrigerators with best performance, this study used refrigerator data collected real time via Internet Things, which were screened to include only effective parameters related compressor and properties (based their practical significance research background) cleaned by applying longitudinal dimensionality reduction transverse reduction. Then, basis an idealized model for data,...

Large Vision-Language Model (LVLM) has seen burgeoning development and increasing attention recently. In this paper, we propose a novel framework, camo-perceptive vision-language framework (CPVLF), to explore whether LVLM can generalize the challenging camouflaged object detection (COD) scenario in training-free manner. During process of generalization, find that due hallucination issues within LVLM, it erroneously perceive objects scenes, producing counterfactual concepts. Moreover, as is...

Retrieving similar trajectories aims to search for the that are close a query trajectory in spatio-temporal domain from large dataset. This is critical variety of applications, like transportation planning and mobility analysis. Unlike previous studies perform retrieval on fine-grained GPS data or single cellular carrier, we investigate feasibility finding multiple carriers, which provide more comprehensive coverage population space. To handle issues spatial bias coarse granularity,...

Objective To evaluate the clinical effect of a novel computer-assisted navigation technique for intraoperative correction femoral rotation deformity in diaphyseal fractures. Methods From November 2015 to 2016, system (BrainLAB, Germany) was used antegrade in-tramedullary nailing 13 patients with shaft fracture intraoperatively restore normal length and fractured femur. They were 11 men 2 women, an average age 38.2 years. The injury affected left side 5 cases right 8. According...