The complex layout of the airport surface, coupled with interrelated vehicle behaviors and densely mixed traffic flows, frequently leads to operational conflict risks. To address this issue, research was conducted on recognition characteristics risk assessment for surface operations in flows. Firstly, a topological network model established based analysis physical structure features surface. Based Monte Carlo simulation method, framework proposed, enabling flows involving aircraft vehicles....

Synthetic accessibility assessment is the pivotal link bridging conceptual phase of design to practical execution synthesis in energetic materials. As chemical space be explored becomes...

As an information carrier, face images contain abundant sensitive information. Due to its natural weak privacy, direct publishing may divulge privacy. Anonymization Technology and Data Encryption are limited by the background knowledge attack means of attackers, which cannot completely content needs image privacy protection. Therefore, this paper proposes a SWP (sliding window publication) algorithm, satisfies differential Firstly, translates gray matrix into one‐dimensional ordered data...

To improve the correct detection ratio of existing universal methods for image steganography, a new steganalysis method based on wavelet package transform (WPT) is presented. Firstly, decompose into three scales through WPT to obtain 85 coefficient subbands together, and extract multi-order absolute characteristic function moments histogram from them as features. And then, normalize these features combine 255-D feature vector each image. Lastly, according this vector, back-propagation (BP)...

Face images, as an information carrier, are rich in sensitive information. Direct publication of these images would cause privacy leak, due to their natural weak privacy. Most the existing protection methods for face adopt data under a non-interactive framework. However, E-effect this framework covers entire image, such that noise influence is uniform across image. To solve problem, paper proposes region growing (RGP), algorithm interactive differential This innovative combines technique...

The privacy protection for face images aims to prevent attackers from accurately identifying target persons through recognition. Inspired by goal-driven reasoning (reverse reasoning), this paper designs a algorithm of local sensitive areas in multiface (face areas) under the interactive framework recognition algorithm, regional growth, and differential privacy. designed named (PPSA), is realized following manner: Firstly, multitask cascaded convolutional network (MTCNN) was adopted recognize...

Sounding source localization is a challenging cross-modal task due to the difficulty of alignment. Although supervised methods achieve encouraging performance, heavy manual annotations are expensive and inefficient. Thus it valuable meaningful develop unsupervised solutions. In this paper, we propose an **U**nsupervised **S**ounding **P**ixel **L**earning (USPL) approach which enables pixel-level sounding in paradigm. We first design mask augmentation based multi-instance contrastive...

Multiplicative noise, including dropout, is widely used to regularize deep neural networks (DNNs), and shown be effective in a wide range of architectures tasks. From an information perspective, we consider injecting multiplicative noise into DNN as training the network solve task with noisy pathways, which leads observation that tends increase correlation between features, so signal-to-noise ratio pathways. However, high feature undesirable, it increases redundancy representations. In this...

Despite recent breakthroughs in reinforcement learning (RL) and imitation (IL), existing algorithms fail to generalize beyond the training environments. In reality, humans can adapt new tasks quickly by leveraging prior knowledge about world such as language descriptions. To facilitate research on language-guided agents with domain adaption, we propose a novel zero-shot compositional policy task, where environments are characterized composition of different attributes. Since there no public...

Face images, as an information carrier, are naturally weak in privacy. If they collected and analyzed by malicious third parties, personal privacy will leak, many other unmeasurable losses occur. Differential protection of face images is mainly being studied under non-interactive frameworks. However, the ε-effect impacts entire image these Besides, noise influence uniform across protected image, during realization Laplace mechanism. The differential interactive mechanisms can protect...

This study aims to develop an intelligent pedestrian traffic light system that utilizes artificial intelligence and visual learning technology optimize the management of lights, in order improve efficiency, reduce congestion, incidence accidents, provide people with a safer more convenient travel environment. The research team uses high-definition cameras capture pedestrians vehicles real-time, computer vision for vehicle detection feature extraction. Through age classification extraction,...

Due to the progress and innovation of technology economy, network is gradually deeply integrated with social life work learning, which promotes profound changes in people's cognition mode life. As an important part field education, piano teaching needs keep pace development, realize development technology, innovate educational methods. The purpose this study establish a diversified education model, improve knowledge, quality, promote high-quality methods, enhance enthusiasm interest...

Real-time tracking of vehicles is important for monitoring whether the roads are congested or not. How to achieve and maintain a high frame rate key problem be solved in practical applications. In recent years, SORT, BOT-SORT [1] other state-of-the-art target algorithms have achieved fruitful results tasks. The author combines YOLOv7(You Only Look Once v7) with bytetrack compares some popular like Deepsort real-time demonstrate that YOLOv7-bytetrack more suitable detection, vehicles.

As the advantage of lower latency and energy consumption, fog computing has been seen as an important enabler to meet harsh requirements Industrial Internet Things (IIoT), which links multiple industrial wireless sensors (IWSs) with limited ability, via offloading computation analysis tasks Fog Access Points (F-APs). However, F-APs often transmit all information-packages IWS in traditional methods, making long delay transmitting massive data-sets. To overcome this issue, we propose a...

This paper proposes the singular general model (SGM) realization approach for two-dimensional (2-D) systems by using elementary operations. The new operation proposed recently co-authors Roesser of regular case will be generalized to system case. Specifically, problem is first formulated as an a well-defined polynomial matrix associated with given transfer function. Then, constructive procedure established based on operations, which guarantees SGM arbitrarily function 2-D systems. Two...