Tarek M. Hamdani
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
Research Subjects
Tarek M. Hamdani has a total of 92 co-authors. Here are the top co-authors by number of shared publications:
- Adel Mohamed Alimi
REsearch Groups in Intelligent Machines (REGIM Lab), National Engineering School of Sfax (ENIS), University of Sfax, Sfax, Tunisia
62 shared publications - Islem Jarraya
REGIM-Lab.: REsearch Groups in Intelligent Machines, University of Sfax, National Engineering School of Sfax (ENIS), Sfax, Tunisia
8 shared publications - Habib Chabchoub
Al Ain University of Science and Technology
7 shared publications Ajith Abraham
Machines Intelligence Research Labs (MIR Labs), Scientific Network for Innovation and Research Excellence Auburn, WA 98071, USA
6 shared publicationsRaja Fdhila
REGIM: REsearch Group on Intelligent Machines, University of Sfax, National School of Engineers (ENIS), BP 1173, 3038, Tunisia
6 shared publications- Thameur Dhieb
REGIM-Lab.: REsearch Groups in Intelligent Machines, University of Sfax, National Engineering School of Sfax (ENIS), Sfax, Tunisia
6 shared publications
Latest Publications
- 29 July 2026
- Published by Springer Nature in Arabian Journal for Science and Engineering
- Vol. 51 (21) , 23873-23903
Green AI Meets Explainability: Beta-Wavelet Convolutional Neural Networks for Image Classification
- 29 July 2026
- Published by Springer Nature in Arabian Journal for Science and Engineering
- Vol. 51 (21) , 23873-23903
Green AI Meets Explainability: Beta-Wavelet Convolutional Neural Networks for Image Classification
- 1 May 2026
- Published by Springer Nature
- p. 95-108
Face Generation from Arabic Text Using GAN-CLS and AraBERT
- 1 May 2026
- Published by Springer Nature
- p. 95-108
Face Generation from Arabic Text Using GAN-CLS and AraBERT
- 2 March 2026
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 85 (3) , 249
Hand-Drawn Image (HDI) dataset: Deep approach for essential tremor recognition
- 2 March 2026
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 85 (3) , 249
Hand-Drawn Image (HDI) dataset: Deep approach for essential tremor recognition
- Open Access
- 21 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 12837-12863
An Integrated Pipeline for Text-Driven 3D Face Synthesis and Enhancement Using Generative Image Models and 3D Reconstruction
This paper presents a modular pipeline for text-driven 3D face synthesis and enhancement that integrates state-of-the-art text-to-image generation, face restoration, and single-image 2D-to-3D reconstruction. The system transforms free-form natural-language prompts into high-fidelity, animatable 3D face meshes through three sequential stages: 1) text-to-image synthesis using Imagen, DALL $\cdot $ E 2, and Stable Diffusion XL; 2) fidelity-controlled face enhancement using CodeFormer; and 3) robust 3D reconstruction using DECA. All models are executed strictly in inference-only mode, without any fine-tuning or retraining, and all results are computed in a paired before/after configuration. Comprehensive 2D evaluations across Blind/Referenceless Image Spatial Quality Evaluator (BRISQUE), Natural Image Quality Evaluator (NIQE), Perception based Image Quality Evaluator (PIQE), Learned Perceptual Image Patch Similarity (LPIPS), and normalized root-mean-square error (NRMSE) and identity similarity show that enhancement improves naturalness and local realism while maintaining geometric stability. Imagen exhibits the most consistent improvements across all metrics. Reference datasets (CelebA, FFHQ, VGGFace2, LFW, CASIA-WebFace) are employed to establish distributional baselines. In 3D, the enhanced pipeline yields refined texture quality (BRISQUE 51.79 to 45.52), strong identity preservation (ArcFace 0.9273), and minimal geometric drift (Chamfer 0.0112). To contextualize absolute geometric fidelity, we further compare reconstructed meshes against ground-truth 3D scans from FRGCv2, reporting an average Chamfer Distance of 0.0187 for the final enhanced reconstruction. The system also supports mask-guided localized edits (e.g., glasses or jewelry removal) without requiring re-generation. Results demonstrate that enhancement significantly stabilizes 3D reconstruction and reduces artifact propagation, validating the modular design. The pipeline offers a practical, extensible foundation for text-driven 3D asset creation in virtual avatars, digital entertainment, and immersive environments.
- 21 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 12837-12863
Open AccessAn Integrated Pipeline for Text-Driven 3D Face Synthesis and Enhancement Using Generative Image Models and 3D Reconstruction
This paper presents a modular pipeline for text-driven 3D face synthesis and enhancement that integrates state-of-the-art text-to-image generation, face restoration, and single-image 2D-to-3D reconstruction. The system transforms free-form natural-language prompts into high-fidelity, animatable 3D face meshes through three sequential stages: 1) text-to-image synthesis using Imagen, DALL $\cdot $ E 2, and Stable Diffusion XL; 2) fidelity-controlled face enhancement using CodeFormer; and 3) robust 3D reconstruction using DECA. All models are executed strictly in inference-only mode, without any fine-tuning or retraining, and all results are computed in a paired before/after configuration. Comprehensive 2D evaluations across Blind/Referenceless Image Spatial Quality Evaluator (BRISQUE), Natural Image Quality Evaluator (NIQE), Perception based Image Quality Evaluator (PIQE), Learned Perceptual Image Patch Similarity (LPIPS), and normalized root-mean-square error (NRMSE) and identity similarity show that enhancement improves naturalness and local realism while maintaining geometric stability. Imagen exhibits the most consistent improvements across all metrics. Reference datasets (CelebA, FFHQ, VGGFace2, LFW, CASIA-WebFace) are employed to establish distributional baselines. In 3D, the enhanced pipeline yields refined texture quality (BRISQUE 51.79 to 45.52), strong identity preservation (ArcFace 0.9273), and minimal geometric drift (Chamfer 0.0112). To contextualize absolute geometric fidelity, we further compare reconstructed meshes against ground-truth 3D scans from FRGCv2, reporting an average Chamfer Distance of 0.0187 for the final enhanced reconstruction. The system also supports mask-guided localized edits (e.g., glasses or jewelry removal) without requiring re-generation. Results demonstrate that enhancement significantly stabilizes 3D reconstruction and reduces artifact propagation, validating the modular design. The pipeline offers a practical, extensible foundation for text-driven 3D asset creation in virtual avatars, digital entertainment, and immersive environments.
- 1 January 2026
- Published by Springer Nature
- p. 218-232
Deep Neural Network for Early Parkinson’s Disease Detection from Offline Hand-Drawing
- 1 January 2026
- Published by Springer Nature
- p. 218-232
Deep Neural Network for Early Parkinson’s Disease Detection from Offline Hand-Drawing
- 10 December 2025
- Published by Springer Nature in Arabian Journal for Science and Engineering
- Vol. 51 (5) , 6897-6917
PDGV Dataset: Investigating Deep Learning CNN for Gait-Based Parkinson’s Disease Recognition
- 10 December 2025
- Published by Springer Nature in Arabian Journal for Science and Engineering
- Vol. 51 (5) , 6897-6917
PDGV Dataset: Investigating Deep Learning CNN for Gait-Based Parkinson’s Disease Recognition
- 27 August 2025
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 84 (41) , 49219-49244
Deepfake detection via image watermarking: A generative adversarial model approach with limited data
- 27 August 2025
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 84 (41) , 49219-49244
Deepfake detection via image watermarking: A generative adversarial model approach with limited data
- 12 July 2025
- Published by Springer Nature in International Journal of Machine Learning and Cybernetics
- p. 1-15
TinySiamese network for biometric analysis
- 12 July 2025
- Published by Springer Nature in International Journal of Machine Learning and Cybernetics
- p. 1-15
TinySiamese network for biometric analysis
- 28 March 2025
- Published by Springer Nature
- p. 34-43
Improving Arabic Fake News Detection Across Context-Aware Attention Deep Model Based on Natural Language Processing
- 28 March 2025
- Published by Springer Nature
- p. 34-43
Improving Arabic Fake News Detection Across Context-Aware Attention Deep Model Based on Natural Language Processing
- 31 December 2024
- Published by Springer Nature in International Journal of Machine Learning and Cybernetics
- Vol. 16 (7-8) , 4251-4264
RLGA-FER: reinforcement learning based on genetic algorithm for facial expression recognition enhancing
- 31 December 2024
- Published by Springer Nature in International Journal of Machine Learning and Cybernetics
- Vol. 16 (7-8) , 4251-4264
RLGA-FER: reinforcement learning based on genetic algorithm for facial expression recognition enhancing