A Comparative Evaluation of Open-Source Vision Language Models for Image Captioning

Basit Öğe Kaydı
dc.contributor.authorKabakus, Abdullah Talha
dc.contributor.authorSenturk, Arafat
dc.date.accessioned2026-07-01T11:36:27Z
dc.date.available2026-07-01T11:36:27Z
dc.date.issued2025
dc.departmentDüzce Üniversitesi
dc.description9th International Artificial Intelligence and Data Processing Symposium, IDAP 2025 -- 6 September 2025 through 7 September 2025 -- Malatya -- 215321
dc.description.abstractVision Language Models (VLMs) have emerged as powerful architectures capable of jointly processing visual and textual inputs across a range of multimodal tasks. However, despite their growing adoption, a systematic evaluation of open-source VLMs on standard image captioning benchmarks remains limited. This study presents a comparative evaluation of three widely used open-source VLMs - LLaVA-7B, MiniCPM-V-8B, and Qwen2.5-VL-7B - on the task of image captioning using the Flickr8k dataset. We conduct a zero-shot assessment of each model's ability to generate descriptive captions for natural images and evaluate their outputs using a combination of lexical and semantic similarity metrics, including BLEU, METEOR, ROUGE-L, and BERTScore (F1). The results reveal notable differences in model behavior and output quality: Qwen2.5-VL-7B consistently achieves the highest scores across most metrics, while LLaVA-7B underperforms, likely due to its focus on instruction-following and visual question answering rather than structured caption generation. Despite Qwen2.5-VL-7B's relative advantage, all models fall short of human-level performance, underscoring the limitations of general-purpose VLMs in producing concise, reference-aligned captions without task-specific fine-tuning. Our findings highlight the importance of dataset alignment, prompt design, and semantic-aware evaluation methods in future research on open-source multimodal systems. © 2025 IEEE.
dc.identifier.doi10.1109/IDAP68205.2025.11222357
dc.identifier.isbn979-833158990-5
dc.identifier.scopus2-s2.0-105025032094
dc.identifier.scopusqualityN/A
dc.identifier.urihttps://doi.org/10.1109/IDAP68205.2025.11222357
dc.identifier.urihttps://hdl.handle.net/20.500.12684/23042
dc.indekslendigikaynakScopus
dc.language.isoen
dc.publisherInstitute of Electrical and Electronics Engineers Inc.
dc.relation.ispartof9th International Artificial Intelligence and Data Processing Symposium, IDAP 2025
dc.relation.publicationcategoryKonferans Öğesi - Uluslararası - Kurum Öğretim Elemanı
dc.rightsinfo:eu-repo/semantics/closedAccess
dc.snmzKA_Scopus_20260623
dc.subjectartificial intelligence
dc.subjectcomputer vision
dc.subjectlanguage model
dc.subjectstring similarity
dc.subjectvision language model
dc.titleA Comparative Evaluation of Open-Source Vision Language Models for Image Captioning
dc.typeConference Object

Dosyalar

Koleksiyon

Scopus İndeksli Yayınlar Koleksiyonu