scholarly journals CscGAN: Conditional Scale-Consistent Generation Network for Multi-Level Remote Sensing Image to Map Translation

2021 ◽  
Vol 13 (10) ◽  
pp. 1936
Author(s):  
Yuanyuan Liu ◽  
Wenbin Wang ◽  
Fang Fang ◽  
Lin Zhou ◽  
Chenxing Sun ◽  
...  
Keyword(s):  
Remote Sensing ◽  
Feature Learning ◽  
Limited Capacity ◽  
Multi Scale ◽  
Multiple Datasets ◽  
Map Generation ◽  
Generation Network ◽  
Multi Level ◽  
Different Levels ◽  
Loss Level

Automatic remote sensing (RS) image to map translation is a crucial technology for intelligent tile map generation. Although existing methods based on a generative network (GAN) generated unannotated maps at a single level, they have limited capacity in handling multi-resolution map generation at different levels. To address the problem, we proposed a novel conditional scale-consistent generation network (CscGAN) to simultaneously generate multi-level tile maps from multi-scale RS images, using only a single and unified model. Specifically, the CscGAN first uses the level labels and map annotations as prior conditions to guide hierarchical feature learning with different scales. Then, a multi-scale discriminator and two multi-scale generators are introduced to describe both high-resolution and low-resolution representations, aiming to improve the similarity of generated maps and thus produce high-quality multi-level tile maps. Meanwhile, a level classifier is designed for further exploring the characteristics of tile maps at different levels. Moreover, the CscGAN is optimized by jointly multi-scale adversarial loss, level classification loss, and scale-consistent loss in an end-to-end manner. Extensive experiments on multiple datasets and study areas demonstrate that the CscGAN outperforms the state-of-the-art methods in multi-level map translation, with great robustness and efficiency.

scholarly journals A Quantitative Validation of Multi-Modal Image Fusion and Segmentation for Object Detection and Tracking

2021 ◽  
Vol 13 (12) ◽  
pp. 2364
Author(s):  
Nicholas LaHaye ◽  
Michael J. Garay ◽  
Brian D. Bue ◽  
Hesham El-Askary ◽  
Erik Linstead
Keyword(s):  
Remote Sensing ◽  
Reference Data ◽  
Deep Belief Networks ◽  
Remotely Sensed Data ◽  
Output Data ◽  
Detection And Tracking ◽  
Multiple Datasets ◽  
Global Coverage ◽  
Quantitative Validation ◽  
Different Levels

In previous works, we have shown the efficacy of using Deep Belief Networks, paired with clustering, to identify distinct classes of objects within remotely sensed data via cluster analysis and qualitative analysis of the output data in comparison with reference data. In this paper, we quantitatively validate the methodology against datasets currently being generated and used within the remote sensing community, as well as show the capabilities and benefits of the data fusion methodologies used. The experiments run take the output of our unsupervised fusion and segmentation methodology and map them to various labeled datasets at different levels of global coverage and granularity in order to test our models’ capabilities to represent structure at finer and broader scales, using many different kinds of instrumentation, that can be fused when applicable. In all cases tested, our models show a strong ability to segment the objects within input scenes, use multiple datasets fused together where appropriate to improve results, and, at times, outperform the pre-existing datasets. The success here will allow this methodology to be used within use concrete cases and become the basis for future dynamic object tracking across datasets from various remote sensing instruments.

scholarly journals Scale-related governance challenges in the water–energy–food nexus: toward a diagnostic approach

2020 ◽  
Author(s):  
Claudia Pahl-Wostl ◽  
Philipp Gorris ◽  
Nicolas Jager ◽  
Larissa Koch ◽  
Louis Lebel ◽  
...  
Keyword(s):  
Institutional Design ◽  
The Sustainable Development ◽  
Multi Scale ◽  
Space And Time ◽  
Trade Offs ◽  
Starting Point ◽  
Governance Challenges ◽  
Multi Level ◽  
Multiple Levels ◽  
Different Levels

AbstractThe notion of a water–energy–food (WEF) nexus was introduced to encourage a more holistic perspective on the sustainable development of natural resources. Most attention has been directed at identifying potential synergies and trade-offs among sectors that could be addressed with improved technologies and management. The governance of the WEF nexus more broadly has received comparatively little attention, and the importance of scale in space and time has been largely ignored. Inspired by scholarship on multi-level governance in individual sectors, this paper identifies four scale-related governance challenges in the WEF nexus, namely: (1) scalar fit, which arises when planning and operating procedures work at different levels along the scales of space and time in different sectors; (2) scalar strategies, wherever the levels at which actors have influence and in which action takes place are contested and negotiated; (3) institutional interplay, where rules and norms in different sectors influence each other at different levels; (4) scalar uncertainty, arising out of the complexity of multi-level and multi-scale interactions. The relevance of these four challenges is illustrated with case studies from developed and developing countries. These examples show the importance of considering multiple levels and scales when assessing the likely effectiveness of WEF nexus governance mechanisms or proposals. The cases underline the need to pay close attention to issues of power, contestation, and negotiation, in addition to the analysis of institutional design. Thus, this paper recommends that nexus governance efforts and proposals be scrutinized for scale assumptions. The four identified challenges offer a suitable starting point for diagnosis.

scholarly journals Adaptive Feature Pyramid Network to Predict Crisp Boundaries via NMS Layer and ODS F-Measure Loss Function

Information ◽  
2022 ◽  
Vol 13 (1) ◽  
pp. 32
Author(s):  
Gang Sun ◽  
Hancheng Yu ◽  
Xiangtao Jiang ◽  
Mingkui Feng
Keyword(s):  
Edge Detection ◽  
Loss Function ◽  
State Of The Art ◽  
Cross Entropy ◽  
Post Processing ◽  
Multi Scale ◽  
Feature Pyramid ◽  
Multi Level ◽  
Different Levels ◽  
F Measure

Edge detection is one of the fundamental computer vision tasks. Recent methods for edge detection based on a convolutional neural network (CNN) typically employ the weighted cross-entropy loss. Their predicted results being thick and needing post-processing before calculating the optimal dataset scale (ODS) F-measure for evaluation. To achieve end-to-end training, we propose a non-maximum suppression layer (NMS) to obtain sharp boundaries without the need for post-processing. The ODS F-measure can be calculated based on these sharp boundaries. So, the ODS F-measure loss function is proposed to train the network. Besides, we propose an adaptive multi-level feature pyramid network (AFPN) to better fuse different levels of features. Furthermore, to enrich multi-scale features learned by AFPN, we introduce a pyramid context module (PCM) that includes dilated convolution to extract multi-scale features. Experimental results indicate that the proposed AFPN achieves state-of-the-art performance on the BSDS500 dataset (ODS F-score of 0.837) and the NYUDv2 dataset (ODS F-score of 0.780).

scholarly journals Consideration of Scale in Remote Sensing of Biodiversity

2020 ◽  
pp. 425-447 ◽  
Author(s):  
John A. Gamon ◽  
Ran Wang ◽  
Hamed Gholizadeh ◽  
Brian Zutta ◽  
Phil A. Townsend ◽  
...  
Keyword(s):  
Remote Sensing ◽  
Vegetation Type ◽  
Careful Consideration ◽  
Biodiversity Monitoring ◽  
Biological Organization ◽  
Multi Scale ◽  
Sampling Scale ◽  
Field Campaigns ◽  
The Relationship ◽  
Different Levels

AbstractA coherent and effective remote sensing (RS) contribution to biodiversity monitoring requires careful consideration of scale in all its dimensions, including spatial, temporal, spectral, and angular, along with biodiversity at different levels of biological organization. Recent studies of the relationship between optical diversity (spectral diversity) and biodiversity reveal a scale dependence that can be influenced by the RS methods used, vegetation type, and degree and nature of disturbance. To better understand these issues, we call for multi-scale field campaigns that test the effect of sampling scale, vegetation type, and degree of disturbance on the ability to detect different kinds of biodiversity, along with the development of improved models that incorporate both physical and biological principles as well as ecological and evolutionary theory. One goal of these studies would be to more closely match instrumentation and sampling scales to biological definitions of biodiversity and so improve optical diversity (spectral diversity) as a proxy for biodiversity. The ultimate goal would be to design and implement a truly effective, “scale-aware” global biodiversity monitoring system employing RS methods. Such a system could improve our understanding of the distribution and functional importance of biodiversity and enhance our ability to manage ecosystems for resilience and sustainability in a changing world.

The Logical and Compositional Organization of the Lecture (Based on a Lecture from M. Mamardashvili’s Course “Lectures on Proust”)

2020 ◽  
pp. 72-81
Author(s):  
Aleksandra A. Talanina ◽  
Keyword(s):  
Academic Discourse ◽  
Basic Function ◽  
Literary Language ◽  
Linguistic Features ◽  
Specific Nature ◽  
Teaching Process ◽  
Discourse Studies ◽  
Multi Level ◽  
Spatiotemporal Coordination ◽  
Different Levels

Functional and stylistic studies give us an idea of linguistic features of speech products, thus enabling style identification. These specific features become most recognizable when comparing styles. Discourse studies, on the contrary, are mainly focused on understanding and describing basic factors of creating a form of a literary language (style) and factors that determine the characteristics of speech products in individual situations within a socially significant sphere. This article presents an analysis of the logical and compositional organization of the lecture as a genre of academic discourse, taking a university lecture from M. Mamardashvili’s course on M. Proust as an example. The specific nature of the lecture genre in academic discourse is determined by its basic function in the teaching process implemented in direct dialogue with the audience. The research is based on the thesis that a lecture is an event that can be analysed using the concept of chronotope. The use of this concept beyond the analysis of fiction is relevant since spatiotemporal coordination is mandatory for any speech product, regardless of the sphere it is created in or the functions it performs. The main feature of the lecture chronotope is multi-level organization, since a lecture has its own internal spatiotemporal coordinates. The lecture chronotope is explicated at different levels of the text (compositional, lexical and grammatical), which are interconnected. Considering this, two interconnected frameworks of the lecture – structural and semantic – are singled out; they provide the logical and compositional organization of the material, which is important to ensure students’ understanding.

Turnout in Multi-Level Systems

2017 ◽  
Author(s):  
Sona N. Golder ◽  
Ignacio Lago ◽  
André Blais ◽  
Elisabeth Gidengil ◽  
Thomas Gschwend
Keyword(s):  
National Level ◽  
The Body ◽  
Political Community ◽  
The Political ◽  
European Level ◽  
National Elections ◽  
Political Issues ◽  
European Elections ◽  
Multi Level ◽  
Different Levels

Voters face different incentives to turn out to vote in one electoral arena versus another. Although turnout is lowest in European elections, it is found that the turnout is only slightly lower in regional than in national elections. Standard accounts suggest that the importance of an election, in terms of the policy-making power of the body to be elected, drives variation in turnout across elections at different levels. This chapter argues that this is only part of the story, and that voter attachment to a particular level also matters. Not all voters feel connected to each electoral arena in the same way. Although for some, their identity and the issues they most care about are linked to politics at the national level, for others, the regional or European level may offer the political community and political issues that most resonate with them.

Parties, Voters, and Multi-Level Elections

2017 ◽  
Author(s):  
Sona N. Golder ◽  
Ignacio Lago ◽  
André Blais ◽  
Elisabeth Gidengil ◽  
Thomas Gschwend
Keyword(s):  
Political Parties ◽  
Driving Force ◽  
Party System ◽  
Voting Behaviour ◽  
Electoral Institutions ◽  
Multi Level ◽  
Different Levels

This chapter argues that individual voting behaviour and the strategies chosen by political parties across multiple electoral arenas should be considered jointly. Existing literature points to the importance of an election as a major driving force in voting behaviour, but it is argued that voters and parties may differ in their assessments of the importance of elections at different levels. The chapter discusses how the effect of the importance of an electoral arena, for both voter and party behaviour, will be conditioned by electoral institutions and characteristics of parties and the party system, in addition to individual voter characteristics contributing to it.

Sign in / Sign up

Export Citation Format

Share Document