User-Guided Data Expansion Modeling to Train Deep Neural Networks With Little Supervision

Article Geochemistry & Geophysics

Semantic Segmentation of Remote Sensing Images With Sparse Annotations

Yuansheng Hua et al.

Summary: This study introduces a framework for semantic segmentation of aerial images based on incomplete annotations, aiming to reduce labeling costs by utilizing sparse scribbled annotations. The proposed method, utilizing the FEature and Spatial relaTional regulArization (FESTA) technique, effectively improves the performance of semantic segmentation.

IEEE GEOSCIENCE AND REMOTE SENSING LETTERS (2022)

添加到收藏夹

Article Geochemistry & Geophysics

Learning CNN Filters From User-Drawn Image Markers for Coconut-Tree Image Classification

Italos Estilon de Souza et al.

Summary: Identifying tree species in aerial images is crucial for land-use classification, plantation monitoring, and disaster impact assessment. Traditional manual identification methods are laborious, costly, and error-prone, necessitating the need for automatic classification methods. We propose a method that minimizes the training images required for a CNN's feature extractor by using user-selected images and learning filters from user-drawn markers, providing better user control and understanding. The method demonstrates advantages over popular CNN models in binary classification of coconut-tree aerial images.

IEEE GEOSCIENCE AND REMOTE SENSING LETTERS (2022)

添加到收藏夹

Article Geochemistry & Geophysics

Multitask Semantic Boundary Awareness Network for Remote Sensing Image Segmentation

Aijin Li et al.

Summary: This study proposes a semantic boundary awareness network (SBANet) to extract boundary information of land cover in high-resolution remote sensing images. The SBANet utilizes a boundary attention module and adaptive weights of multitask learning to capture and learn boundary information. Experimental results demonstrate the effectiveness of SBANet on 2-D semantic labeling datasets.

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING (2022)

添加到收藏夹

Article Computer Science, Artificial Intelligence

Deep High-Resolution Representation Learning for Visual Recognition

Jingdong Wang et al.

Summary: The High-Resolution Network (HRNet) maintains high-resolution representations and exchanges information across resolutions, resulting in superior performance in various applications such as human pose estimation, semantic segmentation, and object detection.

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE (2021)

添加到收藏夹

Review Geography, Physical

Remote sensing image segmentation advances: A meta-analysis

Ioannis Kotaridis et al.

Summary: This article presents conceptual characteristics of image segmentation methods, with a special focus on semantic segmentation. A meta-analysis was conducted through a comprehensive review of recent image segmentation case studies. The paper also addresses interesting subjects such as identification of current problems, image segmentation on non-traditional data, and hot topics for future research.

ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING (2021)

添加到收藏夹

Article Computer Science, Information Systems

High-Resolution Remote Sensing Image Segmentation Framework Based on Attention Mechanism and Adaptive Weighting

Yifan Liu et al.

Summary: The paper introduces an Adaptive Multi-Scale Module (AMSM) and Adaptive Fuse Module (AFM) to address the challenges of semantic segmentation, achieving good segmentation performance through effective fusion of different branch structures and noise filtering.

ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION (2021)

添加到收藏夹

Article Geochemistry & Geophysics

LANet: Local Attention Embedding to Improve the Semantic Segmentation of Remote Sensing Images

Lei Ding et al.

Summary: Two proposed modules, Patch Attention Module (PAM) and Attention Embedding Module (AEM), enhance feature representation in remote sensing images by bridging the gap between high-level and low-level features. Experimental results show that integrating these modules into a baseline fully convolutional network greatly improves performance and outperforms other attention-based methods.

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING (2021)

添加到收藏夹

Article Environmental Sciences