Editorial for the Special Issue - Machine Learning in Computer Vision and Image Sensing: Theory and Applications - Editors’ introduction

Aug 2, 2026

Upload by Aurora

The editorial introduces a Special Issue compiling high-quality research on advanced machine learning models for computer vision and image sensing, spanning both theory and practical applications. It highlights the role of ML in multimodal analysis and outlines accepted contributions across domains such as medical imaging, earth observation, and human activity or detection. The issue includes research on improving annotation efficiency via non-deep active learning, creating the ROSEBUD dataset for river obstacle segmentation in robotic SLAM, and advancing Alzheimer’s diagnosis from MRI using transfer learning and optimization techniques, reporting strong classification accuracy.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages5

Reading Time13 min

AccessFree to download

Editorial for the Special Issue

Scope and motivation

Overview of accepted contributions

Contribution examples (active learning, ROSEBUD dataset, Alzheimer’s classification)

sensors Editorial Editorial for the Special Issue “Machine Learning in Computer Vision and Image Sensing: Theory and Applications” Subrata Chakraborty 1,2,3, * and Biswajeet Pradhan 2,4 Citation: Chakraborty, S.; Pradhan, B. Editorial for the Special Issue“Machine Learning in Computer Vision and Image Sensing: Theory and Applications”. Sensors 2024, 24, 2874 . [https://doi.org/10.3390/s24092874](https://doi.org/10.3390/s24092874) Received: 1 March 2024 Accepted: 25 March 2024 Published: 30 April 2024 Copyright: © 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license ([https://](https://)[ ](https://)[creativecommons.org/licenses/by/](creativecommons.org/licenses/by/)[ ](creativecommons.org/licenses/by/)[4.0/](4.0/)) . 1 School of Science and Technology, University of New England, Armidale, NSW 2351, Australia 2 Centre for Advanced Modelling and Geospatial Information Systems (CAMGIS), School of Civil and Environmental Engineering, Faculty of Engineering & IT, University of Technology Sydney, Sydney, NSW 2007, Australia; [biswajeet.pradhan@uts.edu.au](biswajeet.pradhan@uts.edu.au) 3 Griffith Business School, Griffith University, Nathan, QLD 4111, Australia 4 Earth Observation Centre, Institute of Climate Change, Universiti Kebangsaan Malaysia (UKM), Bangi 43600, Selangor, Malaysia * Correspondence: [subrata.chakraborty@une.edu.au](subrata.chakraborty@une.edu.au) Machine learning (ML) models have experienced remarkable growth in their application for multimodal data analysis over the past decade [1] . The diverse applications of ML models span domains such as medical image [2–4] and signal processing [5,6], remote sensing for earth observation and monitoring [7–9], the detection of daily human activities [10,11], and many more. ML models play a significant role in supporting computer vision and image-sensing applications, helping to unravel complex and real-world challenges. Recent developments in ML empower us to better analyse image and sensor data, motivating extensive research initiatives aimed at addressing applied challenges in multiple domains, including healthcare, agriculture, defence, remote sensing, earth observation, and autonomous navigation. This Special Issue aims to compile a compendium of high-quality research addressing the broad challenges in both the theoretical and applied aspects of advanced ML models in the field of computer vision and image sensing. The 11 papers accepted in this Special Issue encompass key original research in areas such as medical applications, earth observation, and human detection, alongside comprehensive review papers on the application of ML in imaging and sensing. Contribution 1 developed non-deep active learning models capable of significantly improving the annotation efficiency of unlabelled images. While conventional deep neural network-based approaches often require a large number of computation nodes and extensive computation time when selecting the most informative unlabelled images, the proposed method first trains a task model on labelled images to predict unlabelled ones. An uncertainty indicator is then generated for each unlabelled image, with images exhibiting a high uncertainty index nominated for annotation due to their information richness. The proposed method outperforms the current SoTA method by 1% accuracy on CIFAR-10 . Contribution 2 developed the unique River Obstacle Segmentation En-Route by USV Dataset (ROSEBUD), which is assessable for public use in robotic SLAM applications to map both water and non-water objects and obstructions in fluvial images from the water level. ROSEBUD provides an exciting baseline dataset applicable for complex surface navigation through intricate fluvial scenes. The dataset comprises 549 diverse images, including variations in water quality, seasons, and obstacle

FAQs

What is the goal of this Special Issue editorial?

It presents the purpose and scope of a Special Issue focused on advanced machine learning models for computer vision and image sensing, covering both theoretical and applied challenges.

What kinds of contributions does the Special Issue include?

The issue includes original research papers and comprehensive review papers, with work addressing areas such as medical applications, earth observation, and human detection.

What does Contribution 1 focus on?

It develops non-deep active learning models that improve annotation efficiency by selecting the most informative unlabeled images using an uncertainty indicator.

Aug 2, 2026

Upload by Aurora