Extracting roads from satellite imagery is a promising approach to update the dynamic changes of road networks efficiently and timely. However, it is challenging due to the occlusions caused by other objects and the complex traffic environment, the pixel-based methods often generate fragmented roads and fail to predict topological correctness. In this paper, motivated by the road shapes and connections in the graph network, we propose a connectivity attention network (CoANet) to jointly learn the segmentation and pair-wise dependencies. Since the strip convolution is more aligned with the shape of roads, which are long-span, narrow, and distributed continuously. We develop a strip convolution module (SCM) that leverages four strip convolutions to capture long-range context information from different directions and avoid interference from irrelevant regions. Besides, considering the occlusions in road regions caused by buildings and trees, a connectivity attention module (CoA) is proposed to explore the relationship between neighboring pixels. The CoA module incorporates the graphical information and enables the connectivity of roads are better preserved. Extensive experiments on the popular benchmarks (SpaceNet and DeepGlobe datasets) demonstrate that our proposed CoANet establishes new state-of-the-art results. The source code will be made publicly available at: https://mmcheng.net/coanet/.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1109/TIP.2021.3117076 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Naa15 Haploinsufficiency and De Novo Missense Variants Associate With Neurodevelopmental Disorders and Interfere With Neurogenesis and Neuron Development.
Autism Res
January 2025
Center for Medical Genetics and Hunan key Laboratory of Medical Genetics, MOE Key Laboratory of Rare Pediatric Disease, School of Life Sciences, Central South University, Changsha, Hunan, China.
Neurodevelopmental disorders (NDDs) encompass a group of conditions that impact brain development and function, exhibiting significant genetic and clinical heterogeneity. NAA15, the auxiliary subunit of the N-terminal acetyltransferase complex, has garnered attention due to its association with NDDs. However, the precise role of NAA15 in cortical development and its contribution to NDDs remain elusive.
View Article and Find Full Text PDFCommon and distinct neural underpinnings of the association between childhood maltreatment and depression and aggressive behavior.
BMC Psychiatry
January 2025
College of Artificial Intelligence, Southwest University, Chongqing, China.
Background: Although childhood maltreatment (CM) is widely recognized as a transdiagnostic risk factor for various internalizing and externalizing psychological disorders, the neural basis underlying this association remain unclear. The potential reasons for the inconsistent findings may be attributed to the involvement of both common and specific neural pathways that mediate the influence of childhood maltreatment on the emergence of psychopathological conditions.
Methods: This study aimed to delineate both the common and distinct neural pathways linking childhood maltreatment to depression and aggression.
InCoLoTransNet: An Involution-Convolution and Locality Attention-Aware Transformer for Precise Colorectal Polyp Segmentation in GI Images.
J Imaging Inform Med
January 2025
Faculty of Medicine and Pharmacy of Rabat, Mohammed V University of Rabat, Rabat, 10000, Morocco.
Gastrointestinal (GI) disease examination presents significant challenges to doctors due to the intricate structure of the human digestive system. Colonoscopy and wireless capsule endoscopy are the most commonly used tools for GI examination. However, the large amount of data generated by these technologies requires the expertise and intervention of doctors for disease identification, making manual analysis a very time-consuming task.
View Article and Find Full Text PDFTemporal logic inference for interpretable fault diagnosis of bearings via sparse and structured neural attention.
ISA Trans
January 2025
State Key Laboratory of Mechanical System and Vibration, Shanghai Jiao Tong University, Shanghai 200240, China. Electronic address:
This paper addresses the critical challenge of interpretability in machine learning methods for machine fault diagnosis by introducing a novel ad hoc interpretable neural network structure called Sparse Temporal Logic Network (STLN). STLN conceptualizes network neurons as logical propositions and constructs formal connections between them using specified logical operators, which can be articulated and understood as a formal language called Weighted Signal Temporal Logic. The network includes a basic word network using wavelet kernels to extract intelligible features, a transformer encoder with sparse and structured neural attention to locate informative signal segments relevant to decision-making, and a logic network to synthesize a coherent language for fault explanation.
View Article and Find Full Text PDFNeural Rewiring of Resilience: The Effects of Combat Deployment on Functional Network Architecture.
Biol Psychiatry Cogn Neurosci Neuroimaging
January 2025
School of Psychological Sciences, Sagol School of Neuroscience, Tel-Aviv University.
Background: Although combat-deployed soldiers are at a high risk for developing trauma-related psychopathology, most will remain resilient for the duration and aftermath of their deployment tour. The neural basis of this type of resilience is largely unknown, and few longitudinal studies exist on neural adaptation to combat in resilient individuals for whom a pre-exposure measurement was collected. Here, we delineate changes in the architecture of functional brain networks from pre- to post-combat in psychopathology-free, resilient participants.
View Article and Find Full Text PDFWant AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!