Elamur Research⌕
☰
home / papers / 2026 / ImageEval 2026: Culturally Grounded Arabic Multimodal Evaluation

ImageEval 2026: Culturally Grounded Arabic Multimodal Evaluation

Summary

ImageEval 2026 is a shared task that tests whether multimodal models handle Arab cultural content, not just generic scenes. It has two parts. AynVQA covers English and MSA. In Task 1a, a model hears a spoken question with three spoken options and picks one using the image. In Task 1b, a model labels three statements about an image as true or false; only one is visually supported, and the other two are culturally plausible distractors. CRAI-Bench (Task 2) asks systems to predict human cultural-accuracy scores for images generated from Qatari scene captions. Task 1 data comes from OASIS and M2CQA. Task 2 uses 40 verified reference images with five captions each. Fourteen teams competed, and twelve wrote system papers. There were three main findings:

  • Speech: spoken VQA is clearly harder in MSA, largely because of Arabic ASR.
  • Hallucination: choosing the grounded statement jointly is more consistent than checking each statement separately.
  • Cultural scoring: automated scoring can be gamed by caption structure.

The work matters because answer accuracy alone hides weak visual grounding and benchmark shortcuts.

Key points

  • Task 1 splits:

    • Training, development and dev-test sets have 3,000, 500 and 500 examples from OASIS, with voice-cloned speech.
    • The test set has 1,000 examples from M2CQA, with human recordings covering 13 countries.
    • Labels follow a taxonomy of nine cultural categories and 31 subcategories.

  • The Task 1b metric is contrastive instability: CI = 1 − N_consistent / N_partial. It counts triplets with some correct answers that are not fully correct. Baselines were zero-shot Qwen2.5-Omni-3B (1a) and Qwen2.5-VL-3B (1b).

  • The best MSA spoken-VQA accuracy was 8.7 points below the best English result. In Task 1b, a zero-shot system ranked first for MSA.

  • The CRAI dimensions and weights are:

    • CEA 0.30
    • CC 0.20
    • CS 0.20
    • CI 0.20
    • HP −0.10

    Task 2 has 200 gpt-image-1 images. Systems are ranked by Spearman ρ with MAE as tiebreaker, and data is split by reference image (24/8/8).

  • Every Task 2 submission beat the GPT-4o judge. However, the top entry relied mostly on a caption-version prior rather than image evidence.

Interesting ideas

  • I1. A consistency metric scores a grouped set of related items as a whole, so partial luck on a set counts as failure rather than credit. — For Mova (learn): Score evaluation sets in clusters, for example all paraphrases or negations of one fact. This would show where Mova's reasoning is unstable, not just where it is wrong.
  • I2. Each multiple-choice item is turned into one true statement plus two plausible but unsupported ones, which yields hard negatives cheaply. — For Mova (answer): Fable and tale QA could pair each true claim about the story world with believable false claims. This would test whether answers come from the built world rather than from word associations.
  • I3. Selecting the one grounded option jointly beat verifying each option independently, because the task structure becomes part of the decision. — For Mova (answer): When test items have exactly one correct choice, compare all candidate frames against the world together instead of thresholding each in isolation.
  • I4. The winning Task 2 system exploited caption-version metadata, which shows how a benchmark can leak through structure. — For Mova (learn): Add a "blind" negative-control baseline that sees only surface or metadata features. If it scores well, the evaluation set has shortcuts to remove.
  • I5. Splitting data by source item (the reference image) prevents near-duplicates from leaking between train and test. Separately, the gap between synthetic training audio and human test audio exposed a domain shift. — For Mova (learn): Split silver data by source document, and keep a human-written gold test set. This would reveal how much of Mova's training gains are artifacts of the silver data.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.2608.30475 · arXiv 2608.30475

@misc{abdaljalil2026imageeval,
  title = {{ImageEval 2026: Culturally Grounded Arabic Multimodal Evaluation}},
  author = {Abdaljalil, Samir and Bhatti, Hunzalah Hassan and Bashiti, Ahlam and Amir, Farina and Hasan, Md Arid and Mousi, Basel and Durrani, Nadir and Dalvi, Fahim and Ali, Zien Sheikh and Serpedin, Erchin and Kurban, Hasan and Jarrar, Mustafa and Chowdhury, Shammur Absar and Alam, Firoj},
  year = {2026},
  doi = {10.48550/arxiv.2608.30475},
  eprint = {2608.30475},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2608.30475}
}

References

60 references, 6 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Shahad Al-Khalifa, Nadir Durrani, Hend Al-Khalifa, and Firoj Alam. 2025. The landscape of arabic large language models. Communications of the ACM, 68(10):54–61. find
  2. Firoj Alam, Md Arid Hasan, and Shammur Absar Chowdhury. 2025a. SpokenNativQA: Multilingual everyday spoken queries for llms. In Proceedings of the 26th Interspeech Conference (Interspeech 2025), Rotterdam, The Netherlands. ISCA. Firoj Alam, Ali Ezzat Shahroor, Md. find
  3. Arid Hasan, Zien Sheikh Ali, Hunzalah Hassan Bhatti, Mohamed Bayan Kmainasi, Shammur Absar Chowdhury, Basel Mousi, Fahim Dalvi, Nadir Durrani, and Natasa Milic-Frayling. 2025b. OASIS: A multilingual and multimodal dataset for culturally grounded spoken visual qa. arXiv preprint arXiv:2510.06371. ↗ our card
  4. Aisha Alansari and Hamzah Luqman. 2025. AraHalluEval: A fine-grained hallucination evaluation framework for arabic llms. In Proceedings of The Third Arabic Natural Language Processing Conference, pages 148–161. find
  5. Nouar AlDahoul and Yasir Zaki. 2026. NYUAD at ImageEval shared tasks: A multimodal llm for audio-visual question answering and hallucination detection in english and modern standard arabic. In Proceedings of the Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. find
  6. Zien Sheikh Ali, Hamdy Mubarak, Soon-Gyo Jung, Hunzalah Hassan Bhatti, Firoj Alam, and Shammur Absar Chowdhury. 2026. WASIL: In-the-wild arabic spoken interactions with llms. In Proceedings of Interspeech 2026, Sydney, Australia. find our card
  7. Lulwah AlKulaib and Ahmed Jlidi. 2026. DATALabKU at ImageEval 2026 shared tasks: Metric-aware cultural evaluation for CRAI-Bench. In Proceedings of the Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. find
  8. Nada Almarwani, Samah Aloufi, Sakhar B. Alkhereyf, Manal Alhassoun, Manal Almutery, Nouf Alshalawi, and Abdulmohsen Al-Thubaity. 2025. King-domGlimpses: Evaluating saudi cultural representation through text-to-image models. IEEE Access, 13:177822–177845. find
  9. Sultan Alrowili, Younes Samih, Abed Alhakim Freihat, and Mathan Kumar Eswaran. 2026. AraVQA: Building a new Arabic factoid visual question answering dataset from Wikipedia. In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 2026–2042, San Diego, California, United States. Association for Computational Linguistics. find
  10. Fakhraddin Alwajih, Gagan Bhatia, and Muhammad Abdul-Mageed. 2024. Dallah: A dialect-aware multimodal large language model for Arabic. In Proceedings of the Second Arabic Natural Language Processing Conference, pages 320–336, Bangkok, Thailand. Association for Computational Linguistics. find
  11. Fakhraddin Alwajih, Samar M. Magdy, Abdellah El Mekki, Omer Nacar, Youssef Nafea, Safaa Taher Abdelfadil, Abdulfattah Mohammed Yahya, Hamzah Luqman, Nada Almarwani, Samah Aloufi, Baraah Qawasmeh, Houdaifa Atou, Serry Sibaee, Hamzah A. Alsayadi, Walid Al-Dhabyani, Maged S. Al-shaibani, Aya El aatar, Nour Qandos, Rahaf Alhamouri, and 18 others. 2025. Pearl: A multimodal culturally-aware Arabic instruction dataset. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 23048–23079, Suzhou, China. Association for Computational Linguistics. find
  12. Zahra Bayramli, Ayhan Suleymanzade, Na Min An, Huzama Ahmad, Eunsu Kim, Junyeong Park, James Thorne, and Alice Oh. 2025. Diffusion models through a global lens: Are they culturally inclusive? In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 31137–31155, Vienna, Austria. Association for Computational Linguistics. find
  13. Assaf Ben-Kish, Moran Yanuka, Morris Alper, Raja Giryes, and Hadar Averbuch-Elor. 2024. Mitigating open-vocabulary caption hallucinations. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 22680–22698, Miami, Florida, USA. Association for Computational Linguistics. find
  14. Parth Bhalerao, Mounika Yalamarty, Brian Trinh, and Oana Ignat. 2026. When cultures meet: Multicultural text-to-image generation. In Findings of the Association for Computational Linguistics: ACL 2026, pages 35808–35828, San Diego, California, United States. Association for Computational Linguistics. find
  15. Xuweiyi Chen, Ziqiao Ma, Xuejun Zhang, Sihan Xu, Shengyi Qian, Jianing Yang, David Fouhey, and Joyce Chai. 2024. Multi-object hallucination in vision language models. In Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2024, NeurIPS 2024, Vancouver, BC, Canada, December 10 - 15, 2024. find
  16. Zhiyuan Chen, Yuecong Min, Jie Zhang, Bei Yan, Jiahao Wang, Xiaozhen Wang, and Shiguang Shan. 2026. A survey of multimodal hallucination evaluation and detection. International Journal of Computer Vision, 134(3):131. find
  17. Amartya Roy Chowdhury, Tonmoy Rajkhowa, and Sanjeev Sharma. 2025. Towards multilingual spoken visual question answering system using cross-attention. In Proceedings of the 31st International Conference on Computational Linguistics, pages 9165–9175, Abu Dhabi, UAE. Association for Computational Linguistics. find
  18. Wala Elsharif, Marco Agus, Mahmoud Alzubaidi, and James She. 2024. Cultural relevance index: Measuring cultural relevance in ai-generated images. In 2024 IEEE 7th International Conference on Multimedia Information Processing and Retrieval (MIPR), pages 410–416. IEEE. find
  19. Wala Elsharif, Mahmoud Alzubaidi, James She, and Marco Agus. 2025. Ara-Pic: A framework for enhancing arabic cultural representation in ai-generated images. In 2025 IEEE International Conference on Multimedia and Expo Workshops (ICMEW), pages 1–6. IEEE. find
  20. Sara Ghaboura, Ahmed Heakl, Omkar Thawakar, Ali Husain Salem Abdulla Alharthi, Ines Riahi, Abduljalil Radman, Jorma Laaksonen, Fahad Shahbaz Khan, Salman Khan, and Rao Muhammad Anwer. 2025. CAMEL-bench: A comprehensive Arabic LMM benchmark. In Findings of the Association for Computational Linguistics: NAACL 2025, pages 1970–1980, Albuquerque, New Mexico. Association for Computational Linguistics. find
  21. Yash Goyal, Tejas Khot, Douglas Summers-Stay, Dhruv Batra, and Devi Parikh. 2017. Making the V in VQA matter: Elevating the role of image understanding in visual question answering. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 6904–6913. find our card
  22. Ahmed Eid Hassan, Mohamed Eid Abd El-Maguid, Ahmed Hossam Rashad, Nour Eldeen Hossam Mahmoud, Mohamed Hamdy Mohamed, and Abdelrhman Mahmoud Fawzy. 2026. Digilians at ImageEval 2026 task 1a: Modular speech parsing and joint multimodal reasoning for arabic spoken vqa. In Proceedings of the Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. find
  23. Syed Mohaiminul Hoque and Md Sakhawat Hossain. 2026. Team Tokenizers at ImageEval 2026 shared tasks: Answer-first contrastive grounding with qlora. In Proceedings of the Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. find
  24. Drew A. Hudson and Christopher D. Manning. 2019. GQA: A new dataset for real-world visual reasoning and compositional question answering. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 6700– 6709. find our card
  25. Nari Johnson, Deepthi Sudharsan, Hamna, Samantha Dalal, Theo Holroyd, Anja Thieme, Hoda Heidari, Daniela Massiceti, Jennifer Wortman Vaughan, and Cecily Morrison. 2026. Evaluating ai-generated images of cultural artifacts with community-informed rubrics. In The 2026 ACM Conference on Fairness, Accountability, and Transparency, pages 714–774. find
  26. Karima Kadaoui, Hanin Atwany, Hamdan Al-Ali, Abdelrahman Mohamed, Ali Mekky, Sergei Tilga, Natalia Fedorova, Ekaterina Artemova, Hanan Aldarmaki, and Yova Kementchedjhieva. 2026. JEEM: Vision-language understanding in four Arabic dialects. In Findings of the Association for Computational Linguistics: EACL 2026, pages 331–354, Rabat, Morocco. Association for Computational Linguistics. find
  27. Sarah M. Kamel, Shimaa I. Hassan, and Lamiaa Elrefaei. 2023. VAQA: Visual arabic question answering. Arabian Journal for Science and Engineering, 48:10803–10823. find
  28. Nithish Kannen, Arif Ahmad, Marco Andreetto, Vinodkumar Prabhakaran, Utsav Prabhu, Adji Bousso Dieng, Pushpak Bhattacharyya, and Shachi Dave. 2024. Beyond aesthetics: Cultural competence in text-to-image models. In Advances in Neural Information Processing Systems, volume 37, pages 13716–13747. find
  29. Chia-Hsuan Lee, Szu-Lin Wu, Chi-Liang Liu, and Hung yi Lee. 2018. Spoken SQuAD: A Study of Mitigating the Impact of Speech Recognition Errors on Listening Comprehension. In Interspeech 2018, pages 3459–3463. find
  30. Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Dollár, and C. Lawrence Zitnick. 2014. Microsoft coco: Common objects in context. In Computer Vision – ECCV 2014, pages 740–755, Cham. Springer International Publishing. find our card
  31. Holy Lovenia, Wenliang Dai, Samuel Cahyawijaya, Ziwei Ji, and Pascale Fung. 2024. Negative object presence evaluation (NOPE) to measure object hallucination in vision-language models. In Proceedings of the 3rd Workshop on Advances in Language and Vision Research (ALVR), pages 37–58, Bangkok, Thailand. Association for Computational Linguistics. find
  32. Pan Ma, Junlin Liu, and Jun Liu. 2026. Team Falcons at ImageEval 2026 shared tasks: Structure-constrained QLoRA fine-tuning and dual-model consensus verification. In Proceedings of the Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. find
  33. Sina Malakouti, Boqing Gong, and Adriana Kovashka. 2026. Culture in action: Evaluating text-to-image models through social activities. In The Fourteenth International Conference on Learning Representations. find
  34. Ahmed Ayman Ahmed Ezzat Mohamed. 2026. Ahmed Ayman at ImageEval 2026 shared tasks: Measurement-first systems for spoken VQA and cultural accuracy evaluation. In Proceedings of the Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. find
  35. Basel Mousi, Fahim Dalvi, Shammur Chowdhury, Firoj Alam, and Nadir Durrani. 2026a. Said aloud, read different: Cross-modal instability in multimodal models. In Proceedings of Interspeech 2026, Sydney, Australia. Accepted. find
  36. Basel Mousi, Fahim Dalvi, Shammur Absar Chowdhury, Firoj Alam, and Nadir Durrani. 2026b. Once correct, still wrong: Counterfactual hallucination in multilingual vision-language models. In Findings of the Association for Computational Linguistics: ACL 2026, pages 4763–4788, San Diego, California, United States. Association for Computational Linguistics. find
  37. Shravan Nayak, Mehar Bhatia, Xiaofeng Zhang, Verena Rieser, Lisa Anne Hendricks, Sjoerd van Steenkiste, Yash Goyal, Karolina Stanczak, and Aishwarya Agrawal. 2025. CulturalFrames: Assessing cultural expectation alignment in text-to-image models and evaluation metrics. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 20918–20953, Suzhou, China. Association for Computational Linguistics. find
  38. Shravan Nayak, Kanishk Jain, Rabiul Awal, Siva Reddy, Sjoerd Van Steenkiste, Lisa Anne Hendricks, Karolina Stanczak, and Aishwarya Agrawal. 2024. Benchmarking vision language models for cultural understanding. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 5769–5790, Miami, Florida, USA. Association for Computational Linguistics. find
  39. Maher Pasa, Maria Hanini, Amro Najjar, and Hasan Alkhder. 2026. alkhder at ImageEval 2026 shared tasks: Constrained selection and rotation-averaged logit scoring for multimodal hallucination detection. In Proceedings of the Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. find
  40. Jonas Pfeiffer, Gregor Geigle, Aishwarya Kamath, Jan-Martin O. Steitz, Stefan Roth, Ivan Vulić, and Iryna Gurevych. 2022. xGQA: Cross-lingual visual question answering. In Findings of the Association for Computational Linguistics: ACL 2022, pages 2497– 2511, Dublin, Ireland. Association for Computational Linguistics. find
  41. Han Qiu, Jiaxing Huang, Peng Gao, Qin Qi, Xiaoqin Zhang, Ling Shao, and Shijian Lu. 2024. Long-HalQA: Long-context hallucination evaluation for multimodal large language models. arXiv preprint arXiv:2410.09962. ↗
  42. Tonmoy Rajkhowa, Amartya Roy Chowdhury, Sankalp Nagaonkar, Achyut Mani Tripathi, and Mahadeva Prasanna. 2024. TM-PATHVQA: 90000+ Textless Multilingual Questions for Medical Visual Question Answering. In Interspeech 2024, pages 4034–4038. find
  43. Anna Rohrbach, Lisa Anne Hendricks, Kaylee Burns, Trevor Darrell, and Kate Saenko. 2018. Object hallucination in image captioning. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 4035–4045, Brussels, Belgium. Association for Computational Linguistics. find our card
  44. David Romero, Chenyang Lyu, Haryo Akbarianto Wibowo, Teresa Lynn, Injy Hamed, Aditya Nanda Kishore, Aishik Mandal, Alina Dragonetti, Artem Abzaliev, Atnafu Lambebo Tonja, Bontu Fufa Balcha, Chenxi Whitehouse, Christian Salamea, Dan John Velasco, David Ifeoluwa Adelani, David Le Meur, Emilio Villa-Cueva, Fajri Koto, Fauzan Farooqui, and 57 others. 2024. Cvqa: culturally-diverse multilingual visual question answering benchmark. In Proceedings of the 38th International Conference on Neural Information Processing Systems, NIPS ’24, Red Hook, NY, USA. Curran Associates Inc. Md. find
  45. Ashraful Islam Semon and Jihadul Islam. 2026. CUET_InferX at ImageEval 2026 shared tasks: Audio-visual LoRA adaptation of Qwen2.5-Omni for culturally-grounded spoken VQA. In Proceedings of the Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. Md. find
  46. Faisal Sheikh. 2026. md_faisal at ImageEval 2026 shared tasks: Forced-choice reformulation for hallucination detection and a version-prior baseline for cultural relevance. In Proceedings of the Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. find
  47. Norawit Urailertprasert, Peerat Limkonchotiwat, Supasorn Suwajanakorn, and Sarana Nutanong. 2024. SEA-VQA: Southeast asian cultural context dataset for visual question answering. In Proceedings of the 3rd Workshop on Advances in Language and Vision Research (ALVR), pages 173–185, Bangkok, Thailand. Association for Computational Linguistics. find
  48. Viacheslav Vasilev, Julia Agafonova, Nikolai Gerasimenko, Alexander Kapitanov, Polina Mikhailova, Evelina Mironova, and Denis Dimitrov. 2025. Rus-Code: Russian cultural code benchmark for text-to-image generation. In Findings of the Association for Computational Linguistics: NAACL 2025, pages 7656–7672, Albuquerque, New Mexico. Association for Computational Linguistics. find
  49. Andrés Villa, Juan León Alcázar, Alvaro Soto, and Bernard Ghanem. 2025. Behind the magic, merlim: Multi-modal evaluation benchmark for large image-language models. In 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), pages 492–502. find
  50. Dingdong WANG, Junan Li, Jincenzi Wu, Dongchao Yang, Xueyuan Chen, Tianhua Zhang, and Helen M. Meng. 2026. MMSU: A massive multi-task spoken language understanding and reasoning benchmark. In The Fourteenth International Conference on Learning Representations. find
  51. Junyang Wang, Yuhang Wang, Guohai Xu, Jing Zhang, Yukai Gu, Haitao Jia, Jiaqi Wang, Haiyang Xu, Ming Yan, Ji Zhang, and Jitao Sang. 2023. AM-BER: An LLM-free multi-dimensional benchmark for MLLMs hallucination evaluation. arXiv preprint arXiv:2311.07397. ↗
  52. Lei Wang, Jiabang He, Shenshen Li, Ning Liu, and Ee-Peng Lim. 2024. Mitigating fine-grained hallucination by fine-tuning large vision-language models with caption rewrites. In MultiMedia Modeling: 30th International Conference, MMM 2024, Amsterdam, The Netherlands, January 29 – February 2, 2024, Proceedings, Part IV, page 32–45, Berlin, Heidelberg. Springer-Verlag. find
  53. Xiyang Wu, Tianrui Guan, Dianqi Li, Shuaiyi Huang, Xiaoyu Liu, Xijun Wang, Ruiqi Xian, Abhinav Shrivastava, Furong Huang, Jordan Lee Boyd-Graber, Tianyi Zhou, and Dinesh Manocha. 2024. AutoHallusion: Automatic generation of hallucination benchmarks for vision-language models. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 8395–8419, Miami, Florida, USA. Association for Computational Linguistics. find
  54. Muhammad Yamin, Sheikh Abdul Wahid, and Ali Arshad. 2026. Dynamos at ImageEval 2026 shared tasks: Metric-aligned selection and cross-family ensembling for vision-language hallucination detection in english and arabic. In Proceedings of the Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. find
  55. Arnav Yayavaram, Siddharth Yayavaram, Simran Khanuja, Michael Saxon, and Graham Neubig. 2026. CAIRE: Cultural attribution of images with retrieval. In Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers), pages 8320– 8338, Rabat, Morocco. Association for Computational Linguistics. find
  56. Chenyu You, Nuo Chen, Fenglin Liu, Shen Ge, Xian Wu, and Yuexian Zou. 2022. End-to-end spoken conversational question answering: Task, dataset and model. In Findings of the Association for Computational Linguistics: NAACL 2022, pages 1219–1232, Seattle, United States. Association for Computational Linguistics. find
  57. Ahmed Younis. 2026. Ahmed Younis at ImageEval 2026 shared tasks: The judge needs a prior for cultural image evaluation. In Proceedings of the Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. find
  58. Sumaiya Zaman and Miftahul Jannat Rishta. 2026. Nile Nexus at ImageEval 2026 shared tasks: Single-choice reformulation and confidence-calibrated refinement for culturally grounded hallucination detection. In Proceedings of the Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. find
  59. Jerrold H. Zar. 2005. Spearman Rank Correlation. John Wiley & Sons, Ltd. find
  60. Bohan Zhai, Shijia Yang, Chenfeng Xu, Sheng Shen, Kurt Keutzer, Chunyuan Li, and Manling Li. 2023. HallE-Control: Controlling object hallucination in large multimodal models. arXiv preprint arXiv:2310.01779. ↗

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/abdaljalil-2026-imageeval-2026-culturally-grounded-arabic/