Unlocking language boundaries: AraCLIP - transforming Arabic language and image understanding through cross-lingual models
Engineering Applications of Artificial Intelligence, vol. 151, p. 110577
8 publications on cross-lingual image retrieval, Arabic sign language recognition and Arabic NLP — with abstracts, DOIs and BibTeX for each.
Looking for competition and contest results? See Competitions →
The 4 I would hand someone first. Each one appears again below with its abstract and BibTeX.
Engineering Applications of Artificial Intelligence, vol. 151, p. 110577
Proceedings of The Second Arabic Natural Language Processing Conference, p. 102-110
IEEE Spoken Language Technology Workshop (SLT), p. 1006-1013
arXiv preprint
Engineering Applications of Artificial Intelligence, vol. 151, p. 110577
In the domain of image retrieval, the integration of text and images has been transformative, facilitating models that transcend language barriers. This paper introduces Arabic Contrastive Language-Image Pre-training (AraCLIP), an extension of the CLIP model tailored for Arabic image retrieval. AraCLIP leverages the CLIP architecture, introducing Knowledge Distillation to transfer cross-modal knowledge from a pre-trained English model to an Arabic counterpart.
@article{ALBARHAM2025110577,
title = {Unlocking language boundaries: AraCLIP - transforming Arabic language and image understanding through cross-lingual models},
journal = {Engineering Applications of Artificial Intelligence},
volume = {151},
pages = {110577},
year = {2025},
issn = {0952-1976},
doi = {https://doi.org/10.1016/j.engappai.2025.110577},
url = {https://www.sciencedirect.com/science/article/pii/S0952197625005779},
author = {Muhammad Al-Barham and Imad Afyouni and Khalid Almubarak and Ayad Turky and Ibrahim Abaker Targio Hashem and Ali Bou Nassif and Ismail Shahin and Ashraf Elnagar}
}
Proceedings of The Second Arabic Natural Language Processing Conference, p. 102-110
This paper introduces Arabic Contrastive Language-Image Pre-training (AraCLIP), a model designed for Arabic image retrieval tasks, building upon the Contrastive Language-Image Pre-training (CLIP) architecture. AraCLIP leverages Knowledge Distillation to transfer cross-modal knowledge from English to Arabic, enhancing its ability to understand Arabic text and retrieve relevant images.
@inproceedings{al-barham-etal-2024-araclip,
title = "{A}ra{CLIP}: Cross-Lingual Learning for Effective {A}rabic Image Retrieval",
author = "Al-Barham, Muhammad and Afyouni, Imad and Almubarak, Khalid and Elnagar, Ashraf and Turky, Ayad and Hashem, Ibrahim",
booktitle = "Proceedings of The Second Arabic Natural Language Processing Conference",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.arabicnlp-1.9",
pages = "102--110"
}
International Conference on Emerging Trends and Applications in Artificial Intelligence, p. 125-136
This paper investigates different word embedding techniques for Arabic news article classification.
@inproceedings{khaled2023arabic,
title={Arabic News Articles Classification Using Different Word Embeddings},
author={Khaled, M Moneb and Al-Barham, Muhammad and Alomari, Osama Ahmad and Elnagar, Ashraf},
booktitle={International Conference on Emerging Trends and Applications in Artificial Intelligence},
pages={125--136},
year={2023},
organization={Springer}
}
International Conference on Emerging Trends and Applications in Artificial Intelligence, p. 226-237
This paper presents a transfer learning approach for Arabic Sign Language alphabet classification.
@inproceedings{al2023arabic,
title={Arabic Sign Language Alphabet Classification via Transfer Learning},
author={Al-Barham, Muhammad and Alomari, Osama Ahmad and Elnagar, Ashraf},
booktitle={International Conference on Emerging Trends and Applications in Artificial Intelligence},
pages={226--237},
year={2023},
organization={Springer}
}
IEEE Spoken Language Technology Workshop (SLT), p. 1006-1013
This paper presents MASC, a Massive Arabic Speech Corpus for automatic speech recognition research.
@inproceedings{al2023masc,
title={MASC: Massive Arabic Speech Corpus},
author={Al-Fetyani, Mohammad and Al-Barham, Muhammad and Abandah, Gheith and Alsharkawi, Adham and Dawas, Maha},
booktitle={2022 IEEE Spoken Language Technology Workshop (SLT)},
pages={1006--1013},
year={2023},
organization={IEEE}
}
International Conference on Emerging Trends and Applications in Artificial Intelligence, p. 279-291
This paper proposes using Marine Predatory Algorithm for feature selection in speech emotion recognition tasks.
@inproceedings{alomari2023marine,
title={Marine Predatory Algorithm for Feature Selection in Speech Emotion Recognition},
author={Alomari, Osama Ahmad and Al-Barham, Muhammad and Elnagar, Ashraf},
booktitle={International Conference on Emerging Trends and Applications in Artificial Intelligence},
pages={279--291},
year={2023},
organization={Springer}
}
arXiv preprint
This paper presents a comprehensive RGB dataset for Arabic Alphabets Sign Language recognition.
@article{al2023rgb,
title={RGB Arabic Alphabets Sign Language Dataset},
author={Al-Barham, Muhammad and Alsharkawi, Adham and Al-Yaman, Musa and Al-Fetyani, Mohammad and Elnagar, Ashraf and SaAleek, Ahmad Abu and Al-Odat, Mohammad},
journal={arXiv preprint arXiv:2301.11932},
year={2023}
}
13th International Conference on Information and Communication Systems (ICICS), p. 226-231
This paper presents deep learning approaches for Arabic Sign Language recognition.
@inproceedings{al2022arabic,
title={Arabic Sign Language Recognition Using Deep Learning Models},
author={Al-Barham, Muhammad and Sa'Aleek, Ahmad Abu and Al-Odat, Mohammad and Hamad, Ghada and Al-Yaman, Musa and Elnagar, Ashraf},
booktitle={2022 13th International Conference on Information and Communication Systems (ICICS)},
pages={226--231},
year={2022},
organization={IEEE}
}