Bassem Bsir
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Bassem Bsir has a total of 6 co-authors. Here are the top co-authors by number of shared publications:
- Mounir Zrigui
University of Monastir
5 shared publications - Fathi Kallel
Laboratory of Advanced Technologies for Image & Signal Processing, National School of Electronics and Telecommunications, University of Sfax, Sfax, Tunisia
1 shared publication - Imene Sekkiou
Laboratory of Electromagnetism and Guided Optics, Department of Electrical Engineering, Faculty of Science and Technology, University of Abdelhamid Ibn Badis, Mostaganem, Algeria
1 shared publication - Marwa Kahia
ATMS Lab, Advanced Technologies for Medicine and Signals, ENIS, Sfax University, Sfax, Tunisia, Tunisia
1 shared publication Mounir Zrigui
University of Monastir
1 shared publication- Nabil Khoufi
University of Sfax
1 shared publication
Latest Publications
- Open Access
- 23 January 2025
- Published by MDPI AG in Preprints.org
Modified Multi-Head Attention Transformer (MMHAT) for Skin Image Classification
Skin cancer detection is a crucial task in dermatology, with accurate classification of skin lesions playing a key role in improving patient outcomes. This paper proposes a modified architecture based on Multi-Head Attention Transformer (MMHAT) for skin image classification, specifically targeting the detection of malignant melanoma and melanocytic nevus. The Transformer model leverages self-attention mechanisms to capture complex spatial relationships and contextual information across different regions of skin images, which is essential for distinguishing between various types of lesions. Unlike conventional Convolutional Neural Networks (CNNs), which are limited by kernel size and layer depth, the Transformer’s self-attention mechanism enables it to capture long-range dependencies and reduce reliance on external information, allowing for better internal data correlation. In this work, we utilize a multi-head self-attention module to extract image features, which allows the model to focus on multiple aspects of the image simultaneously, enhancing its ability to detect subtle patterns often missed by traditional CNNs. The model is evaluated on the publicly available ISIC 2018 dataset, containing a diverse range of dermoscopic images. Experimental results show that the proposed Transformer model achieves an accuracy of 0.91, outperforming conventional models such as ResNet50 and Inception V3, which achieved accuracy of 0.90 and 0.84, respectively. Furthermore, experiments with and without multi-head attention demonstrate the significant impact of the attention mechanism in improving model performance, particularly in skin cancer detection.
- 23 January 2025
- Published by MDPI AG in Preprints.org
Open AccessModified Multi-Head Attention Transformer (MMHAT) for Skin Image Classification
Skin cancer detection is a crucial task in dermatology, with accurate classification of skin lesions playing a key role in improving patient outcomes. This paper proposes a modified architecture based on Multi-Head Attention Transformer (MMHAT) for skin image classification, specifically targeting the detection of malignant melanoma and melanocytic nevus. The Transformer model leverages self-attention mechanisms to capture complex spatial relationships and contextual information across different regions of skin images, which is essential for distinguishing between various types of lesions. Unlike conventional Convolutional Neural Networks (CNNs), which are limited by kernel size and layer depth, the Transformer’s self-attention mechanism enables it to capture long-range dependencies and reduce reliance on external information, allowing for better internal data correlation. In this work, we utilize a multi-head self-attention module to extract image features, which allows the model to focus on multiple aspects of the image simultaneously, enhancing its ability to detect subtle patterns often missed by traditional CNNs. The model is evaluated on the publicly available ISIC 2018 dataset, containing a diverse range of dermoscopic images. Experimental results show that the proposed Transformer model achieves an accuracy of 0.91, outperforming conventional models such as ResNet50 and Inception V3, which achieved accuracy of 0.90 and 0.84, respectively. Furthermore, experiments with and without multi-head attention demonstrate the significant impact of the attention mechanism in improving model performance, particularly in skin cancer detection.
- 25 July 2024
- Published by Springer Nature
- p. 380-388
Deep Learning Based Transformers for Keyword Extraction
- 25 July 2024
- Published by Springer Nature
- p. 380-388
Deep Learning Based Transformers for Keyword Extraction
- Open Access
- 31 January 2024
- Published by Slovenian Association Informatika in Informatica
- Vol. 48 (1)
Prediction of Author’s Profile basing on Fine-Tuning BERT model
The task of author profiling consists in specifying the infer-demographic features’ of the social networks’ users by studying their published content or the interactions between them. In the literature, many research works were conducted to enhance the accuracy of the techniques used in this process. In fact, the existing methods can be divided into two types: simple linear mod-els and complex deep neural network models. Among them, the transformer-based model exhibited the highest efficiency in NLP analysis in several lan-guages (English, German, French, Turk, Arabic, etc.). Despite their good per-formance, these approaches do not cover author profiling analysis and, thus, should be further enhanced. So, we propose in this paper a new deep learning strategy by training a customized transformer-model to learn the optimal fea-tures of our dataset. In this direction, we fine-tune the model by using the trans-fer learning approach to improve the results with random initialization. We have achieved about 79% of accuracy by modifying model to apply the retrain-ing process using PAN 2018 authorship dataset.
- 31 January 2024
- Published by Slovenian Association Informatika in Informatica
- Vol. 48 (1)
Open AccessPrediction of Author’s Profile basing on Fine-Tuning BERT model
The task of author profiling consists in specifying the infer-demographic features’ of the social networks’ users by studying their published content or the interactions between them. In the literature, many research works were conducted to enhance the accuracy of the techniques used in this process. In fact, the existing methods can be divided into two types: simple linear mod-els and complex deep neural network models. Among them, the transformer-based model exhibited the highest efficiency in NLP analysis in several lan-guages (English, German, French, Turk, Arabic, etc.). Despite their good per-formance, these approaches do not cover author profiling analysis and, thus, should be further enhanced. So, we propose in this paper a new deep learning strategy by training a customized transformer-model to learn the optimal fea-tures of our dataset. In this direction, we fine-tune the model by using the trans-fer learning approach to improve the results with random initialization. We have achieved about 79% of accuracy by modifying model to apply the retrain-ing process using PAN 2018 authorship dataset.
- 16 May 2019
- Published by Springer Nature in Lecture Notes in Computer Science
- p. 621-631
Document Model with Attention Bidirectional Recurrent Network for Gender Identification
- 16 May 2019
- Published by Springer Nature in Lecture Notes in Computer Science
- p. 621-631
Document Model with Attention Bidirectional Recurrent Network for Gender Identification
- 14 April 2019
- Published by Springer Nature
- p. 792-800
Gender Identification: A Comparative Study of Deep Learning Architectures
- 14 April 2019
- Published by Springer Nature
- p. 792-800
Gender Identification: A Comparative Study of Deep Learning Architectures
- 30 September 2018
- Published by Instituto Politecnico Nacional/Centro de Investigacion en Computacion in Computación y Sistemas
- Vol. 22 (3) , 757-766
Enhancing Deep Learning Gender Identification with Gated Recurrent Units Architecture in Social Text
Enhancing Deep Learning Gender Identification with Gated Recurrent Units Architecture in Social Text
- 30 September 2018
- Published by Instituto Politecnico Nacional/Centro de Investigacion en Computacion in Computación y Sistemas
- Vol. 22 (3) , 757-766
Enhancing Deep Learning Gender Identification with Gated Recurrent Units Architecture in Social Text
Enhancing Deep Learning Gender Identification with Gated Recurrent Units Architecture in Social Text
- 8 August 2018
- Published by Springer Nature
- p. 393-402
Bidirectional LSTM for Author Gender Identification
- 8 August 2018
- Published by Springer Nature
- p. 393-402
Bidirectional LSTM for Author Gender Identification