A Study of Commonsense Reasoning over Visual Object Properties

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Kolari, Abhishek, Khojasteh, Mohammadhossein, Jiang, Yifan, Hengst, Floris den, Ilievski, Filip
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914256508682240
author Kolari, Abhishek
Khojasteh, Mohammadhossein
Jiang, Yifan
Hengst, Floris den
Ilievski, Filip
author_facet Kolari, Abhishek
Khojasteh, Mohammadhossein
Jiang, Yifan
Hengst, Floris den
Ilievski, Filip
contents Inspired by human categorization, object property reasoning involves identifying and recognizing low-level details and higher-level abstractions. While current visual question answering (VQA) studies consider multiple object properties, such as size, they typically blend perception and reasoning and lack representativeness in terms of reasoning and image categories, making it unclear whether and how vision-language models (VLMs) abstract and reason over depicted objects. To this end, we introduce a systematic evaluation framework comprising images of three representative types, three reasoning levels of increasing complexity, and four object property dimensions, informed by prior work on common sense. We develop a procedure to instantiate this framework in two VQA object reasoning benchmarks: OPTICS-CNT, comprising 360 images paired with 1,080 multi-level, count-based questions, and OPTICS-CMP, with 2.1k comparison questions. Experiments with 12 state-of-the-art VLMs in zero-shot settings reveal significant limitations relative to humans, with the best-performing model achieving below 40% counting and 70% comparison accuracy. VLMs struggle particularly with photographic images, counterfactual reasoning, physical and functional properties, and higher counts. We make the OPTICS benchmark data and code available to support future work on scalable benchmarking methods, generalized annotation guidelines, and advanced reasoning VLMs.
format Preprint
id arxiv_https___arxiv_org_abs_2508_10956
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Study of Commonsense Reasoning over Visual Object Properties
Kolari, Abhishek
Khojasteh, Mohammadhossein
Jiang, Yifan
Hengst, Floris den
Ilievski, Filip
Computer Vision and Pattern Recognition
Artificial Intelligence
Inspired by human categorization, object property reasoning involves identifying and recognizing low-level details and higher-level abstractions. While current visual question answering (VQA) studies consider multiple object properties, such as size, they typically blend perception and reasoning and lack representativeness in terms of reasoning and image categories, making it unclear whether and how vision-language models (VLMs) abstract and reason over depicted objects. To this end, we introduce a systematic evaluation framework comprising images of three representative types, three reasoning levels of increasing complexity, and four object property dimensions, informed by prior work on common sense. We develop a procedure to instantiate this framework in two VQA object reasoning benchmarks: OPTICS-CNT, comprising 360 images paired with 1,080 multi-level, count-based questions, and OPTICS-CMP, with 2.1k comparison questions. Experiments with 12 state-of-the-art VLMs in zero-shot settings reveal significant limitations relative to humans, with the best-performing model achieving below 40% counting and 70% comparison accuracy. VLMs struggle particularly with photographic images, counterfactual reasoning, physical and functional properties, and higher counts. We make the OPTICS benchmark data and code available to support future work on scalable benchmarking methods, generalized annotation guidelines, and advanced reasoning VLMs.
title A Study of Commonsense Reasoning over Visual Object Properties
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2508.10956