scispace - formally typeset
Search or ask a question
Topic

Bounding overwatch

About: Bounding overwatch is a research topic. Over the lifetime, 966 publications have been published within this topic receiving 15156 citations.


Papers
More filters
Posted Content
TL;DR: In this article, an attention-based activation map approach is proposed to improve the classification of tiny regions, which also produces locations using feature map scores learned from class labels, achieving state-of-the-art performance.
Abstract: Citrus juices and fruits are commodities with great economic potential in the international market, but productivity losses caused by mites and other pests are still far from being a good mark. Despite the integrated pest mechanical aspect, only a few works on automatic classification have handled images with orange mite characteristics, which means tiny and noisy regions of interest. On the computational side, attention-based models have gained prominence in deep learning research, and, along with weakly supervised learning algorithms, they have improved tasks performed with some label restrictions. In agronomic research of pests and diseases, these techniques can improve classification performance while pointing out the location of mites and insects without specific labels, reducing deep learning development costs related to generating bounding boxes. In this context, this work proposes an attention-based activation map approach developed to improve the classification of tiny regions called Two-Weighted Activation Mapping, which also produces locations using feature map scores learned from class labels. We apply our method in a two-stage network process called Attention-based Multiple Instance Learning Guided by Saliency Maps. We analyze the proposed approach in two challenging datasets, the Citrus Pest Benchmark, which was captured directly in the field using magnifying glasses, and the Insect Pest, a large pest image benchmark. In addition, we evaluate and compare our models with weakly supervised methods, such as Attention-based Deep MIL and WILDCAT. The results show that our classifier is superior to literature methods that use tiny regions in their classification tasks, surpassing them in all scenarios by at least 16 percentage points. Moreover, our approach infers bounding box locations for salient insects, even training without any location labels.

5 citations

Proceedings ArticleDOI
05 Jun 2000
TL;DR: This work presents an effective under-bounding safeguard against system model violations in OBE processing and Simulation examples in state estimation and speech processing demonstrate the efficacy of the under- bounding safeguard.
Abstract: Optimal bounding ellipsoid (OBE) identification algorithms are noted for their simplicity and ability to leverage model error-bound knowledge for improved parameter convergence. However, the OBE convergence rate is dependent on the pointwise "tightness" of the model error-bound estimates. Since the least upper bound on the model error is often unknown, the convergence rate is compromised by the need to overestimate error-bounds lest the integrity of the process be violated by underestimation. We present an effective under-bounding safeguard against system model violations in OBE processing. Simulation examples in state estimation and speech processing demonstrate the efficacy of the under-bounding safeguard.

5 citations

Proceedings ArticleDOI
01 Sep 2017
TL;DR: A novel componentwise bounding procedure applicable to both real and complex nonlinear systems with additive disturbances and is shown to have highly desirable properties, such as being monotone and positive.
Abstract: How to bound the state vector trajectory of a nonlinear system in a way so that the obtained bound be of practical value is an open problem. If some norm is employed for bounding the state vector trajectory, then this norm should be carefully selected and the state vector components suitably scaled. In addition, practical applications usually require separate bounds on every state variable. Bearing this context in mind, we develop a novel componentwise bounding procedure applicable to both real and complex nonlinear systems with additive disturbances. A bound on the magnitude of the evolution of each state variable is obtained by computing a single trajectory of a well-specified “bounding” system constructed from the original system equations and the available disturbance bounds. The bounding system is shown to have highly desirable properties, such as being monotone and positive. We provide preliminary results establishing that key stability features are preserved by the bounding system for systems in triangular form.

5 citations

01 Jan 2008
TL;DR: In this paper, a multiscale analysis method is presented for capturing the dynamics inside fiber-reinforced composites at both the structural scale and the microscopic scale, with a cost that is much less than solving the full micro-scale model over the entire macroscopic domain.
Abstract: The dissertation provides new multiscale methods for the analysis of heterogeneous media. The first part of the dissertation treats heterogeneous media using the theory of linear elasticity. In this context, a methodology is presented for bounding the higher L norms, 2 ¤ p ¤ 8, of the local stress and strain fields inside random elastic media. Optimal lower bounds that are given in terms of the applied loading and the volume (area) fractions for random two-phase composites are presented. These bounds provide a means to measure load transfer across length scales relating the excursions of the local fields to applied loads. The second part of the dissertation treats heterogeneous media using the peridynamic formulation of nonlocal continuum mechanics. In this context, a multiscale analysis method is presented for capturing the dynamics inside fiber-reinforced composites at both the structural scale and the microscopic scale. The method provides a multiscale numerical method with a cost that is much less than solving the full micro-scale model over the entire macroscopic domain.

5 citations


Network Information
Related Topics (5)
Robustness (computer science)
94.7K papers, 1.6M citations
85% related
Optimization problem
96.4K papers, 2.1M citations
85% related
Matrix (mathematics)
105.5K papers, 1.9M citations
82% related
Nonlinear system
208.1K papers, 4M citations
81% related
Artificial neural network
207K papers, 4.5M citations
80% related
Performance
Metrics
No. of papers in the topic in previous years
YearPapers
2023714
20221,629
2021155
202075
201973
201850