{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,24]],"date-time":"2026-06-24T14:54:21Z","timestamp":1782312861262,"version":"3.54.5"},"reference-count":50,"publisher":"Association for Computing Machinery (ACM)","issue":"7","license":[{"start":{"date-parts":[[2026,6,24]],"date-time":"2026-06-24T00:00:00Z","timestamp":1782259200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/legalcode"}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Multimedia Comput. Commun. Appl."],"published-print":{"date-parts":[[2026,7,31]]},"abstract":"<jats:p>The automatic detection of persuasive strategies in advertisements presents a uniquely multimodal challenge at the intersection of vision, language, and social cognition. While recent advances in Large Vision-Language Models (LVLMs) offer promising capabilities for such tasks, current approaches often rely on restrictive evaluation schemes that do not reflect the inherently multi-label nature of persuasive messaging. In this work, we reframe persuasion strategy detection as a genuine multi-label classification problem and propose a principled evaluation framework to enhance interpretability and robustness. We apply this approach to both image and video datasets, examine their characteristics, and introduce novel input-agnostic baselines that achieve macro F1 scores of 0.082 and 0.289 on the respective test sets. As part of this analysis, we study label co-occurrence patterns and dataset ambiguities, providing insights that inform both model interpretation and future dataset design. To assess the native capabilities of LVLMs, we benchmark three open source models\u2014PaliGemma, PaliGemma2, and Qwen2.5-VL\u2014on the image persuasion dataset. Under zero-shot conditions, we demonstrate that querying each strategy individually with a logit-based decision threshold outperforms guided text generation. The best-performing zero-shot model, Qwen2.5-VL, achieves a macro F1 score of 0.227 and a sample F1 score of 0.234 on the image test set, and 0.381 and 0.400 respectively on the video test set. We further explore and compare two lightweight fine-tuning strategies that update only small subsets of model parameters while keeping the remaining weights frozen: fine-tuning of the image-to-text tokens linear projection and Low Rank Adaptation (LoRA) of the language model. Linear projector fine-tuning yields a top macro F1 score of 0.396 on the image test set, marking a substantial improvement over zero-shot performance. To evaluate cross-modal generalization, we apply fine-tuned image models to the video dataset. Our experiments reveal that, while projection-based fine-tuning enables partial knowledge transfer from image to video (macro F1 = 0.416 on a testing subset), LoRA adaptation severely disrupts cross-modal performance (macro F1 = 0.189). Finally, we perform a per-strategy performance analysis, looking into annotator- and data-centric factors that may influence LVLM performance. These findings highlight the viability of open-weight LVLMs for fine-grained persuasion analysis and suggest efficient pathways for domain-specific adaptation under realistic resource constraints.<\/jats:p>","DOI":"10.1145\/3788874","type":"journal-article","created":{"date-parts":[[2026,1,19]],"date-time":"2026-01-19T14:07:04Z","timestamp":1768831624000},"page":"1-27","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":1,"title":["Principled Evaluation of Multi-Label Persuasion in Advertisements with Large Vision-Language Models"],"prefix":"10.1145","volume":"22","author":[{"ORCID":"https:\/\/orcid.org\/0009-0004-2769-9752","authenticated-orcid":false,"given":"Iv\u00e1n","family":"Mart\u00edn-Fern\u00e1ndez","sequence":"first","affiliation":[{"name":"THAU Group, Information Processing and Telecommunications Center, Universidad Polit\u00e9cnica de Madrid, Madrid, Spain"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2312-6672","authenticated-orcid":false,"given":"Mihai Gabriel","family":"Constantin","sequence":"additional","affiliation":[{"name":"AI Multimedia Lab, National University of Science and Technology Politehnica Bucharest, Bucharest, Romania"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0005-0535-1306","authenticated-orcid":false,"given":"Bogdan","family":"Ionescu","sequence":"additional","affiliation":[{"name":"AI Multimedia Lab, National University of Science and Technology Politehnica Bucharest, Bucharest, Romania"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4285-6224","authenticated-orcid":false,"given":"Manuel","family":"Gil-Mart\u00edn","sequence":"additional","affiliation":[{"name":"THAU Group, Information Processing and Telecommunications Center, Universidad Polit\u00e9cnica de Madrid, Madrid, Spain"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3877-0089","authenticated-orcid":false,"given":"Fernando","family":"Fern\u00e1ndez-Mart\u00ednez","sequence":"additional","affiliation":[{"name":"THAU Group, Information Processing and Telecommunications Center, Universidad Polit\u00e9cnica de Madrid, Madrid, Spain"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"320","published-online":{"date-parts":[[2026,6,24]]},"reference":[{"key":"e_1_3_2_2_2","unstructured":"Anthropic. 2024. Claude 3.5 Sonnet Model Card Addendum. Retrieved from https:\/\/www-cdn.anthropic.com\/fed9cc193a14b84131812372d8d5857f8f304c52\/Model_Card_Claude_3_Addendum.pdf"},{"key":"e_1_3_2_3_2","doi-asserted-by":"publisher","DOI":"10.21437\/odyssey.2024-41"},{"key":"e_1_3_2_4_2","unstructured":"Lucas Beyer Andreas Steiner Andr\u00e9 Susano Pinto Alexander Kolesnikov Xiao Wang Daniel Salz Maxim Neumann Ibrahim Alabdulmohsin Michael Tschannen Emanuele Bugliarello et al. 2024. Paligemma: A versatile 3b VLM for transfer. arXiv:2407.07726. Retrieved from https:\/\/arxiv.org\/abs\/2407.07726"},{"key":"e_1_3_2_5_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.emnlp-main.608"},{"key":"e_1_3_2_6_2","unstructured":"Nimet Beyza Bozdag Shuhaib Mehri Xiaocheng Yang Hyeonjeong Ha Zirui Cheng Esin Durmus Jiaxuan You Heng Ji Gokhan Tur and Dilek Hakkani-T\u00fcr. 2025. Must read: A systematic survey of computational persuasion. arXiv:2505.07775. Retrieved from https:\/\/arxiv.org\/abs\/2505.07775"},{"key":"e_1_3_2_7_2","unstructured":"S\u00e9bastien Bubeck Varun Chandrasekaran Ronen Eldan Johannes Gehrke Eric Horvitz Ece Kamar Peter Lee Yin Tat Lee Yuanzhi Li Scott Lundberg et al. 2023. Sparks of artificial general intelligence: Early experiments with gpt-4. arXiv:2303.12712. Retrieved from https:\/\/arxiv.org\/abs\/2303.12712"},{"key":"e_1_3_2_8_2","doi-asserted-by":"publisher","DOI":"10.1109\/WACV61041.2025.00567"},{"issue":"1","key":"e_1_3_2_9_2","first-page":"70","article-title":"The SCIENCE of persuasion","volume":"14","author":"Cialdini Robert B.","year":"2004","unstructured":"Robert B. Cialdini. 2004. The SCIENCE of persuasion. Scientific American Mind 14, 1 (2004), 70\u201377. Retrieved from http:\/\/www.jstor.org\/stable\/24939368","journal-title":"Scientific American Mind"},{"key":"e_1_3_2_10_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D19-1565"},{"issue":"3","key":"e_1_3_2_11_2","first-page":"181","article-title":"A systematic literature review on neuromarketing\u2014A new horizon of marketing","volume":"6","author":"Aradhita Deb.","year":"2021","unstructured":"Aradhita, Deb. 2021. A systematic literature review on neuromarketing\u2014A new horizon of marketing. International Journal of Scientific Development and Research 6, 3 (2021), 181\u2013191.","journal-title":"International Journal of Scientific Development and Research"},{"key":"e_1_3_2_12_2","doi-asserted-by":"publisher","DOI":"10.1017\/pan.2021.15"},{"key":"e_1_3_2_13_2","volume-title":"Proceedings of the Iberian Languages Evaluation Forum (IberLEF \u201924), Co-Located with the 40th Conference of the Spanish Society for Natural Language Processing (SEPLN \u201924)","author":"Esteban-Romero Sergio","year":"2024","unstructured":"Sergio Esteban-Romero, Jaime Bellver-Soler, I. Mart\u0131n-Fern\u00e1ndez, M. Gil-Mart\u0131n, Luis Fernando D\u2019Haro, and F. Fern\u00e1ndez-Mart\u0131nez. 2024. THAU-UPM at EmoSPeech-IberLEF2024: Efficient adaptation of mono-modal and multi-modal large language models for automatic speech emotion recognition. In Proceedings of the Iberian Languages Evaluation Forum (IberLEF \u201924), Co-Located with the 40th Conference of the Spanish Society for Natural Language Processing (SEPLN \u201924). Retrieved from https:\/\/ceur-ws.org\/"},{"key":"e_1_3_2_14_2","doi-asserted-by":"publisher","DOI":"10.3390\/sym17081349"},{"key":"e_1_3_2_15_2","unstructured":"Jiaming Han Renrui Zhang Wenqi Shao Peng Gao Peng Xu Han Xiao Kaipeng Zhang Chris Liu Song Wen Ziyu Guo et al. 2023. ImageBind-LLM: Multi-modality instruction tuning. arXiv:2309.03905. Retrieved from https:\/\/arxiv.org\/abs\/2309.03905"},{"key":"e_1_3_2_16_2","unstructured":"Edward J. Hu Yelong Shen Phillip Wallis Zeyuan Allen-Zhu Yuanzhi Li Shean Wang Lu Wang and Weizhu Chen. 2021. LoRA: Low-rank adaptation of large language models. arXiv:2106.09685. Retrieved from https:\/\/arxiv.org\/abs\/2106.09685"},{"key":"e_1_3_2_17_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.123"},{"key":"e_1_3_2_18_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2014.35"},{"key":"e_1_3_2_19_2","volume-title":"Proceedings of the 12th International Conference on Learning Representations","author":"Khandelwal Ashmit","year":"2024","unstructured":"Ashmit Khandelwal, Aditya Agrawal, Aanisha Bhattacharyya, Yaman Kumar, Somesh Singh, Uttaran Bhattacharya, Ishita Dasgupta, Stefano Petrangeli, Rajiv Ratn Shah, Changyou Chen, and Balaji Krishnamurthy. 2024. Large content and behavior models to understand, simulate, and optimize content and behavior. In Proceedings of the 12th International Conference on Learning Representations. Retrieved from https:\/\/openreview.net\/forum?id=TrKq4Wlwcz"},{"key":"e_1_3_2_20_2","unstructured":"Junseo Kim Jongwook Han Dongmin Choi Jongwook Yoon Eun-Ju Lee and Yohan Jo. 2025. PVP: An image dataset for personalized visual persuasion with persuasion strategies viewer characteristics and persuasiveness ratings. arXiv:2506.00481. Retrieved from https:\/\/arxiv.org\/abs\/2506.00481"},{"key":"e_1_3_2_21_2","unstructured":"Michal Kosinski. 2023. Theory of mind may have spontaneously emerged in large language models. Retrieved from https:\/\/arxiv.org\/vc\/arxiv\/papers\/2302\/2302.02083v1.pdf"},{"key":"e_1_3_2_22_2","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v37i1.25076"},{"key":"e_1_3_2_23_2","doi-asserted-by":"publisher","DOI":"10.1086\/605297"},{"key":"e_1_3_2_24_2","first-page":"19730","volume-title":"Proceedings of the 40th International Conference on Machine Learning, Vol. 202","author":"Li Junnan","year":"2023","unstructured":"Junnan Li, Dongxu Li, Silvio Savarese, and Steven Hoi. 2023. BLIP-2: Bootstrapping language-image pre-training with frozen image encoders and large language models. In Proceedings of the 40th International Conference on Machine Learning, Vol. 202. Andreas Krause, Emma Brunskill, Kyunghyun Cho, Barbara Engelhardt, Sivan Sabato, and Jonathan Scarlett (Eds.), Proceedings of Machine Learning Research, 19730\u201319742. Retrieved from https:\/\/proceedings.mlr.press\/v202\/li23q.html"},{"key":"e_1_3_2_25_2","unstructured":"Haotian Liu Chunyuan Li Yuheng Li and Yong Jae Lee. 2023. Improved baselines with visual instruction tuning. arXiv:2310.03744. Retrieved from https:\/\/arxiv.org\/abs\/2310.03744"},{"key":"e_1_3_2_26_2","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1516"},{"key":"e_1_3_2_27_2","first-page":"1","volume-title":"Proceedings of the 9th Workshop on Argument Mining","author":"Liu Zhexiong","year":"2022","unstructured":"Zhexiong Liu, Meiqi Guo, Yue Dai, and Diane Litman. 2022. ImageArg: A multi-modal tweet dataset for image persuasiveness mining. In Proceedings of the 9th Workshop on Argument Mining. Gabriella Lapesa, Jodi Schneider, Yohan Jo, Sougata, and Saha (Eds.), International Conference on Computational Linguistics, 1\u201318. Retrieved from https:\/\/aclanthology.org\/2022.argmining-1.1\/"},{"key":"e_1_3_2_28_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51701.2025.02172"},{"key":"e_1_3_2_29_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.naacl-long.203"},{"key":"e_1_3_2_30_2","first-page":"4788","volume-title":"Proceedings of the Winter Conference on Applications of Computer Vision (WACV \u201925)","author":"Malakouti Sina","year":"2025","unstructured":"Sina Malakouti, Aysan Aghazadeh, Ashmit Khandelwal, and Adriana Kovashka. 2025. Benchmarking VLMs\u2019 reasoning about persuasive atypical images. In Proceedings of the Winter Conference on Applications of Computer Vision (WACV \u201925). IEEE Computer Society, 4788\u20134798."},{"key":"e_1_3_2_31_2","doi-asserted-by":"publisher","DOI":"10.1145\/3689062.3689083"},{"key":"e_1_3_2_32_2","doi-asserted-by":"publisher","DOI":"10.3390\/s25061661"},{"key":"e_1_3_2_33_2","unstructured":"Meta. 2024. Llama 3.2 model card. Retrieved November 25 2014 from https:\/\/github.com\/meta-llama\/llama-models\/blob\/main\/models\/llama3_2\/MODEL_CARD.md"},{"key":"e_1_3_2_34_2","unstructured":"OpenAI. 2023. GPT-4 technical report. arXiv:2303.08774. Retrieved from https:\/\/arxiv.org\/abs\/2303.08774"},{"key":"e_1_3_2_35_2","doi-asserted-by":"publisher","DOI":"10.1145\/2663204.2663260"},{"key":"e_1_3_2_36_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/N18-1146"},{"key":"e_1_3_2_37_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.chb.2013.02.014"},{"key":"e_1_3_2_38_2","unstructured":"Qwen Team. 2025. Qwen2.5-VL technical report. arXiv:2502.13923. Retrieved from https:\/\/arxiv.org\/abs\/2502.13923"},{"key":"e_1_3_2_39_2","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-23808-6_10"},{"key":"e_1_3_2_40_2","unstructured":"Somesh Singh Yaman K. Singla S. I. Harini and Balaji Krishnamurthy. 2024. Measuring and improving persuasiveness of large language models. arXiv:2410.02653. Retrieved from https:\/\/arxiv.org\/abs\/2410.02653"},{"key":"e_1_3_2_41_2","unstructured":"Andreas Steiner Andr\u00e9 Susano Pinto Michael Tschannen Daniel Keysers Xiao Wang Yonatan Bitton Alexey Gritsenko Matthias Minderer Anthony Sherbondy Shangbang Long et al. 2024. PaliGemma 2: A family of versatile VLMs for transfer. arXiv:2412.03555. Retrieved from https:\/\/arxiv.org\/abs\/2412.03555"},{"key":"e_1_3_2_42_2","doi-asserted-by":"publisher","DOI":"10.1145\/3544548.3581431"},{"key":"e_1_3_2_43_2","doi-asserted-by":"publisher","DOI":"10.1145\/2872427.2883081"},{"key":"e_1_3_2_44_2","unstructured":"Gemini Team Rohan Anil Sebastian Borgeaud Jean-Baptiste Alayrac Jiahui Yu Radu Soricut Johan Schalkwyk Andrew M. Dai Anja Hauth Katie Millican et al. 2023. Gemini: A family of highly capable multimodal models. arXiv:2312.11805. Retrieved from https:\/\/arxiv.org\/abs\/2312.11805"},{"key":"e_1_3_2_45_2","unstructured":"Gemma Team Thomas Mesnard Cassidy Hardin Robert Dadashi Surya Bhupatiraju Shreya Pathak Laurent Sifre Morgane Rivi\u00e8re Mihir Sanjay Kale Juliette Love et al. 2024. Gemma: Open models based on Gemini research and technology. arXiv:2403.08295. Retrieved from https:\/\/arxiv.org\/abs\/2403.08295"},{"key":"e_1_3_2_46_2","unstructured":"Qwen Team. 2024. Qwen2.5: A Party of Foundation Models. Retrieved from https:\/\/qwenlm.github.io\/blog\/qwen2.5\/"},{"key":"e_1_3_2_47_2","first-page":"1498","volume-title":"Proceedings of the International Conference on Recent Advances in Natural Language Processing (RANLP 2021)","author":"Vorakitphan Vorakit","year":"2021","unstructured":"Vorakit Vorakitphan, Elena Cabrio, and Serena Villata. 2021. \u201cDon\u2019t discuss\u201d: Investigating semantic and argumentative features for supervised propagandist message detection and classification. In Proceedings of the International Conference on Recent Advances in Natural Language Processing (RANLP 2021), Ruslan Mitkov and Galia Angelova (Eds.), INCOMA Ltd., 1498\u20131507. Retrieved from https:\/\/aclanthology.org\/2021.ranlp-1.168\/"},{"key":"e_1_3_2_48_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P19-1566"},{"key":"e_1_3_2_49_2","unstructured":"An Yang Baosong Yang Binyuan Hui Bo Zheng Bowen Yu Chang Zhou Chengpeng Li Chengyuan Li Dayiheng Liu Fei Huang et al. 2024. Qwen2 technical report. arXiv:2407.10671. Retrieved from https:\/\/arxiv.org\/abs\/2407.10671"},{"key":"e_1_3_2_50_2","doi-asserted-by":"publisher","unstructured":"Zhao Yukun Liying Xu Zhen Huang Kaiping Peng Martin Seligman Evelyn Li and Feng Yu. 2023. AI chatbot responds to emotional cuing. DOI: 10.31234\/osf.io\/9ymfz","DOI":"10.31234\/osf.io\/9ymfz"},{"key":"e_1_3_2_51_2","unstructured":"Yuze Zhao Jintao Huang Jinghan Hu Xingjun Wang Yunlin Mao Daoze Zhang Zeyinzi Jiang Zhikai Wu Baole Ai Ang Wang Wenmeng Zhou and Yingda Chen. 2024. SWIFT: A scalable lightweight infrastructure for fine-tuning. arXiv:2408.05517. Retrieved from https:\/\/arxiv.org\/abs\/2408.05517"}],"container-title":["ACM Transactions on Multimedia Computing, Communications, and Applications"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3788874","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,24]],"date-time":"2026-06-24T14:44:03Z","timestamp":1782312243000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3788874"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,24]]},"references-count":50,"journal-issue":{"issue":"7","published-print":{"date-parts":[[2026,7,31]]}},"alternative-id":["10.1145\/3788874"],"URL":"https:\/\/doi.org\/10.1145\/3788874","relation":{},"ISSN":["1551-6857","1551-6865"],"issn-type":[{"value":"1551-6857","type":"print"},{"value":"1551-6865","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6,24]]},"assertion":[{"value":"2025-06-19","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-01-04","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-06-24","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}