Deep Temporal Reasoning in Video Language Models: A Cross-Linguistic Evaluation of Action Duration and Completion through Perfect Times
Fuente:
arXiv
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916771925065728 |
|---|---|
| author | Loginova, Olga Loguinova, Sofía Ortega |
| author_facet | Loginova, Olga Loguinova, Sofía Ortega |
| contents | Human perception of events is intrinsically tied to distinguishing between completed (perfect and telic) and ongoing (durative) actions, a process mediated by both linguistic structure and visual cues. In this work, we introduce the \textbf{Perfect Times} dataset, a novel, quadrilingual (English, Italian, Russian, and Japanese) multiple-choice question-answering benchmark designed to assess video-language models (VLMs) on temporal reasoning. By pairing everyday activity videos with event completion labels and perfectivity-tailored distractors, our dataset probes whether models truly comprehend temporal dynamics or merely latch onto superficial markers. Experimental results indicate that state-of-the-art models, despite their success on text-based tasks, struggle to mirror human-like temporal and causal reasoning grounded in video. This study underscores the necessity of integrating deep multimodal cues to capture the nuances of action duration and completion within temporal and causal video dynamics, setting a new standard for evaluating and advancing temporal reasoning in VLMs. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_00928 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Deep Temporal Reasoning in Video Language Models: A Cross-Linguistic Evaluation of Action Duration and Completion through Perfect Times Loginova, Olga Loguinova, Sofía Ortega Computer Vision and Pattern Recognition Computation and Language Human perception of events is intrinsically tied to distinguishing between completed (perfect and telic) and ongoing (durative) actions, a process mediated by both linguistic structure and visual cues. In this work, we introduce the \textbf{Perfect Times} dataset, a novel, quadrilingual (English, Italian, Russian, and Japanese) multiple-choice question-answering benchmark designed to assess video-language models (VLMs) on temporal reasoning. By pairing everyday activity videos with event completion labels and perfectivity-tailored distractors, our dataset probes whether models truly comprehend temporal dynamics or merely latch onto superficial markers. Experimental results indicate that state-of-the-art models, despite their success on text-based tasks, struggle to mirror human-like temporal and causal reasoning grounded in video. This study underscores the necessity of integrating deep multimodal cues to capture the nuances of action duration and completion within temporal and causal video dynamics, setting a new standard for evaluating and advancing temporal reasoning in VLMs. |
| title | Deep Temporal Reasoning in Video Language Models: A Cross-Linguistic Evaluation of Action Duration and Completion through Perfect Times |
| topic | Computer Vision and Pattern Recognition Computation and Language |
| url | https://arxiv.org/abs/2506.00928 |