Abstract We consider medical image transformation problems where a grayscale is transformed into color image. The colorized should have the same features as input because extra synthesized can increase possibility of diagnostic errors. In this paper, to secure images and improve quality images, well leverage unpaired training data, colorization network proposed based on cycle generative adversarial (CycleGAN) model, combining perceptual loss function total variation (TV) function. Visual...

Crowdsensing applications provide platforms for sharing sensing data collected by mobile devices. A blockchain system has the potential to replace a traditional centralized trusted third party crowdsensing services perform operations that involve evaluating quality of data, finishing payment, and storing so forth. The requirements which are codified as smart contracts executed evaluate in blockchain. However, regardless fact may actually be sufficient, one key challenge is malicious...

Crowdsourcing for Human Intelligence Tasks (HIT) has been widely used to crowdsource human knowledge, such as image annotation machine learning. We use a public blockchain play the role of traditional centralized HIT systems, that deals with cryptocurrency payments and acts trustworthy judge resolve disputes between worker requester in decentralized setting, preventing false-reporting free-riding. Our approach neither uses expensive cryptographic tools, zero-knowledge proofs, nor sends...

Machine learning (ML) has led to disruptive innovations in many fields, such as medical diagnoses. A key enabler for ML is large training data, but existing are not fully exploited by because of data silos and privacy concerns. Federated (FL) a promising distributed paradigm address this problem. On the other hand, FL approaches vulnerable poisoning attacks or leakage from malicious aggregator client. This article proposes an auditable scheme with Byzantine robustness against client: The...

On a battlefield, tactical moving objects, such as airplanes and tanks, need to track monitor. It is crucial for headquarters know the event occurrence, especially complex events, of objects in real time, jet fighter invading ship collision. However, radars or other sensors are only able detect primitive simple events objects. challenge manage large amounts incoming analyze their relationships infer time. In this paper, we provide clear solution dealing with challenges by using processing...

Human intelligence tasks (HITs) are widely utilized for crowdsourcing human knowledge, such as labeling images machine learning. Centralized platforms face challenges of a single point failure and lack service transparency. Existing blockchain-based approaches overlook the low scalability problem permissionless blockchains or inconveniently rely on existing ground-truth data root trust to evaluate quality workers' answers. We propose scheme ensuring dual fairness (i.e., preventing...

Complex Event Processing (CEP) over tactical moving objects would detect threats in real time and response defense timely. The input data are stream collected by radar, sonar or other sensors. From the beginning of preprocessing to end threat detection defense, system CEP goes through a series complicated procedures, working based on rules be defined dynamically users. In this paper, we design implement rule-based that allows users insert, update delete rule specifications without code programming.