Khaled Rezeg
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Khaled Rezeg has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
Nadia Ben Seghir
LINFI Laboratory, University of Biskra, Algeria
2 shared publications- Okba Kazar
Department of Computer Science College of Arts Sciences IT & Communication University of Kalba Sharjah United Arab Emirates
2 shared publications - Khalida Zanouda1 shared publication
- Lyna Miloudi
LINFI Laboratory, Department of Computer Science, Mohamed Khider University, Biskra 07000, Algeria
1 shared publication - Mohamed Kotoub Miloudi
Department of Computer Science, Sorbonne Université, 75005 Paris, France
1 shared publication - Mohamed Tayeb Laskri
Department of Computer Science LRI Laboratory Badji Mokhtar University Annaba Algérie
1 shared publication
Latest Publications
- Open Access
- 29 August 2026
- Published by MDPI AG in International Journal of Plant Biology
- Vol. 17 (9) , 80
Image-Based Identification of Crop and Weed Species at the Seedling Stage: Deep Learning Classifiers Rely on Acquisition Context Beyond Plant Morphology
Distinguishing crop from weed species at the seedling stage is a fine-grained morphological discrimination problem. We study it on the public Plant Seedlings benchmark, which contains twelve species (three crops and nine weeds) imaged at the seedling stage. Automated classifiers report near-ceiling accuracy on this benchmark. Yet its images are acquired in controlled trays containing soil, gravel, rulers, barcodes, and printed labels, so a model may identify a species from its acquisition context rather than its morphology. We audit two architectures, EfficientNet-B7 and ViT-B/16, trained on the V2 dataset (5539 images) and probed with plant-only, background-only, and background-swapped inputs. Near-ceiling models (96.1% and 96.4% over three seeds) recover the correct species for up to 47.7% of samples from the background alone (chance 8.3%) and lose about 60 points under background swapping. Context reliance is therefore a property of the benchmark, not any single architecture. Tracing this to its source, an independent learned representation of the plant-free background alone identifies the species at 72.4%. The reliance is correctable end to end: a consistency-regularisation scheme retains 92.1% full-image accuracy for the transformer with no segmentation at inference, at an architecture-dependent cost. Reported accuracy thus partly measures acquisition context, not morphology; morphological grounding should be measured and reported alongside accuracy.
- 29 August 2026
- Published by MDPI AG in International Journal of Plant Biology
- Vol. 17 (9) , 80
Open AccessImage-Based Identification of Crop and Weed Species at the Seedling Stage: Deep Learning Classifiers Rely on Acquisition Context Beyond Plant Morphology
Distinguishing crop from weed species at the seedling stage is a fine-grained morphological discrimination problem. We study it on the public Plant Seedlings benchmark, which contains twelve species (three crops and nine weeds) imaged at the seedling stage. Automated classifiers report near-ceiling accuracy on this benchmark. Yet its images are acquired in controlled trays containing soil, gravel, rulers, barcodes, and printed labels, so a model may identify a species from its acquisition context rather than its morphology. We audit two architectures, EfficientNet-B7 and ViT-B/16, trained on the V2 dataset (5539 images) and probed with plant-only, background-only, and background-swapped inputs. Near-ceiling models (96.1% and 96.4% over three seeds) recover the correct species for up to 47.7% of samples from the background alone (chance 8.3%) and lose about 60 points under background swapping. Context reliance is therefore a property of the benchmark, not any single architecture. Tracing this to its source, an independent learned representation of the plant-free background alone identifies the species at 72.4%. The reliance is correctable end to end: a consistency-regularisation scheme retains 92.1% full-image accuracy for the transformer with no segmentation at inference, at an architecture-dependent cost. Reported accuracy thus partly measures acquisition context, not morphology; morphological grounding should be measured and reported alongside accuracy.
- Open Access
- 24 June 2026
- Published by Atlantis Press SARL
- p. 194-204
From Data-Centric AI to Organizational Intelligence: Why Data Quality, Governance, and Human Oversight Determine the Success of AI Adoption in MENA Organizations
- 24 June 2026
- Published by Atlantis Press SARL
- p. 194-204
Open AccessFrom Data-Centric AI to Organizational Intelligence: Why Data Quality, Governance, and Human Oversight Determine the Success of AI Adoption in MENA Organizations
- 7 January 2021
- Published by Springer Nature in Arabian Journal for Science and Engineering
- Vol. 46 (4) , 3653-3674
A Novel Quranic Search Engine Using an Ontology-Based Semantic Indexing
- 7 January 2021
- Published by Springer Nature in Arabian Journal for Science and Engineering
- Vol. 46 (4) , 3653-3674
A Novel Quranic Search Engine Using an Ontology-Based Semantic Indexing
- 1 January 2018
- Published by IGI Global Scientific Publishing
- p. 308-339
A Personalized Approach for Web Service Discovery in Distributed Environments
Web services are meaningful only if potential users may find and execute them. Universal description discovery and integration (UDDI) help businesses, organizations, and other web services providers to discover and reach to the service(s) by providing the URI of the WSDL file. However, it does not offer a mechanism to choose a web service based on its quality. The standard also lacks sufficient semantic description in the content of web services. This lack makes it difficult to find and compose suitable web services during analysis, search, and matching processes. In addition, a central UDDI suffers from one centralized point problem and the high cost of maintenance. To get around these problems, the authors propose in this chapter a novel framework based on mobile agent and metadata catalogue for web services discovery. Their approach is based on user profile in order to discover appropriate web services, meeting customer requirements in less time and taking into account the QoS properties.
- 1 January 2018
- Published by IGI Global Scientific Publishing
- p. 308-339
A Personalized Approach for Web Service Discovery in Distributed Environments
Web services are meaningful only if potential users may find and execute them. Universal description discovery and integration (UDDI) help businesses, organizations, and other web services providers to discover and reach to the service(s) by providing the URI of the WSDL file. However, it does not offer a mechanism to choose a web service based on its quality. The standard also lacks sufficient semantic description in the content of web services. This lack makes it difficult to find and compose suitable web services during analysis, search, and matching processes. In addition, a central UDDI suffers from one centralized point problem and the high cost of maintenance. To get around these problems, the authors propose in this chapter a novel framework based on mobile agent and metadata catalogue for web services discovery. Their approach is based on user profile in order to discover appropriate web services, meeting customer requirements in less time and taking into account the QoS properties.
- 1 October 2015
- Published by IGI Global Scientific Publishing in International Journal of Information Technology and Web Engineering
- Vol. 10 (4) , 20-43
A Decentralized Framework for Semantic Web Services Discovery Using Mobile Agent
Web services discovery provided by the UDDI registries is relatively primitive. It does not take into account the continuous growth in the number of services on the Web. The UDDI standard has been proposed and used for Web service publication and discovery. However, it does not allow users to choose the best provider. It does not offer a mechanism to choose a Web service based on its quality. The standard also lacks of sufficient semantic description in the content of Web services, this lack makes it difficult to find and compose suitable Web services during analysis, search, and matching processes. In addition, a central UDDI suffers from one centralized point problem and the high cost of maintenance. To get around these problems, the authors propose in this paper a novel framework based on mobile agent and metadata catalogue for Web services discovery. Their approach is based on user profile in order to discover appropriate Web services, meeting customer requirements, in less time and taking into account the QoS properties.
- 1 October 2015
- Published by IGI Global Scientific Publishing in International Journal of Information Technology and Web Engineering
- Vol. 10 (4) , 20-43
A Decentralized Framework for Semantic Web Services Discovery Using Mobile Agent
Web services discovery provided by the UDDI registries is relatively primitive. It does not take into account the continuous growth in the number of services on the Web. The UDDI standard has been proposed and used for Web service publication and discovery. However, it does not allow users to choose the best provider. It does not offer a mechanism to choose a Web service based on its quality. The standard also lacks of sufficient semantic description in the content of Web services, this lack makes it difficult to find and compose suitable Web services during analysis, search, and matching processes. In addition, a central UDDI suffers from one centralized point problem and the high cost of maintenance. To get around these problems, the authors propose in this paper a novel framework based on mobile agent and metadata catalogue for Web services discovery. Their approach is based on user profile in order to discover appropriate Web services, meeting customer requirements, in less time and taking into account the QoS properties.
- Open Access
- 30 April 2010
- Published by AICIT in Journal of Convergence Information Technology
- Vol. 5 (2) , 28-35
Geospatial Web Services Semantic Discovery Approach Using Quality
- 30 April 2010
- Published by AICIT in Journal of Convergence Information Technology
- Vol. 5 (2) , 28-35
Open AccessGeospatial Web Services Semantic Discovery Approach Using Quality