While CNN-based models have made remarkable progress on human pose estimation, what spatial dependencies they capture to localize keypoints remains unclear. In this work, we propose a model called Trans-Pose, which introduces Transformer for estimation. The attention layers built in enable our long-range relationships efficiently and also can reveal the predicted key-points rely on. To predict keypoint heatmaps, last layer acts as an aggregator, collects contributions from image clues forms...

Current state-of-the-art trajectory methods do not perform well in the terminal airspace that surrounds an airport due to its complex structure and frequently changing flight postures of aircraft. Since aircraft takes off or lands must follow a specified procedure, this paper will learn data-driven prediction model from many historical trajectories improve accuracy robustness airspace. A regularization method is utilized reconstruct each obtain high-quality with equal time intervals no...

Air traffic control (ATC) hazard feature extraction is a key information retrieval task for air records. While text-based ranks term importance based solely on statistical results, we aim to use external knowledge extract features that meet the definition of hazards. This paper proposes method expert define and construct analysis framework. We illustrate model training process using communication navigation surveillance (CNS) data, which includes candidate generation, vectorization,...

Learning long-term dependences (LTDs) with recurrent neural networks (RNNs) is challenging due to their limited internal memories. In this paper, we propose a new external memory architecture for RNNs called an addressable and working (EALWM)-augmented RNN. This has two distinct advantages over existing architectures, namely the division of into parts-long-term memory-with both capability learn LTDs without suffering from vanishing gradients necessary assumptions. The experimental results on...

Due to the strong propagation causality of delays between airports, this paper proposes a delay prediction model based on deep graph neural network study from perspective an airport network. We regard airports as nodes and use directed construct airports’ relationship. For adjacent weights edges are measured by spherical distance them, while number flight pairs them is utilized for connected flights. On basis, diffusion convolution kernel constructed capture characteristics it further...

LiDAR-based 3D object detection pushes forward an immense influence on autonomous vehicles. Due to the limitation of intrinsic properties LiDAR, fewer points are collected at objects farther away from sensor. This imbalanced density point clouds degrades accuracy but is generally neglected by previous works. To address challenge, we propose a novel two-stage framework, named SIENet. Specifically, design Spatial Information Enhancement (SIE) module predict spatial shapes foreground within...

Trajectory-based operation is an air traffic control mode with more accuracy, safety, and efficiency, effective measure for future airspace management under the conditions of large flow, high density, short interval, which significantly improves utilization resources. Aircraft trajectory prediction key technology trajectory-based operations, requiring information high-precision to achieve high-density in airspace. The current state-of-the-art forecasting methods, perform low accuracy...

Recent years have seen a dramatic growth of semantic web on the data level, but unfortunately not schema which contains mostly concept hierarchies. The shortage schemas makes difficult to be used in many applications, so learning from becomes an increasingly pressing issue. In this paper we propose novel approach -BelNet, combines description logics (DLs) with Bayesian networks. way BelNet is capable understand and capture semantics one hand, handle incompleteness during procedure other...

The air traffic control sector (ATCS) is the basic unit of airspace system. If we can identify congestion an ATCS, it will help provide decision support for planning and daily operations. However, current methods mainly characterize from static structure dynamic operational features, resulting in poor generalization operability. To this end, propose a deep learning method perspective complex networks. It takes aircraft as nodes to construct network utilizes complexity indices it. So, problem...

Deep contextualized embeddings, as learned by large pre-training models, have proven highly effective in various downstream natural language processing tasks. However, the embedding space these models lacks explicit regularization, leading to underfitting and substantial costs during large-scale training on huge corpora. In this paper, we present a novel approach learning deep introducing linguistic knowledge regularization. Specifically, our proposed model, MorPhEMe (Morphology Phonology...

3D detection task plays a crucial role in the perception system of intelligent vehicles. LiDAR-based detectors perform well on particular autonomous driving benchmarks, but may poorly generalize to other domains. Existing domain adaptive methods usually require annotation-related statistics or continuous refinement pseudo-labels. The former is not always feasible for practical applications, while latter lacks sufficient accurate supervision. In this work, we propose novel unsupervised...

While CNN-based models have made remarkable progress on human pose estimation, what spatial dependencies they capture to localize keypoints remains unclear. In this work, we propose a model called \textbf{TransPose}, which introduces Transformer for estimation. The attention layers built in enable our long-range relationships efficiently and also can reveal the predicted rely on. To predict keypoint heatmaps, last layer acts as an aggregator, collects contributions from image clues forms...

With the rapid development of civil aviation transportation, an increasing number airport groups are formed. However, existing literature on fairness mostly focuses among airlines. There is no research realization scheduling airports with overlapping resources in group. The goal this paper to comprehensively consider efficiency and slot scheduling, where should include both interairline interairport fairness. Subsequently, we developed a collaborative optimization model for group that takes...