A Benchmark for Crime Surveillance Video Analysis with Large Models
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916612928438272 |
|---|---|
| author | Chen, Haoran Yi, Dong Cao, Moyan Huang, Chensen Zhu, Guibo Wang, Jinqiao |
| author_facet | Chen, Haoran Yi, Dong Cao, Moyan Huang, Chensen Zhu, Guibo Wang, Jinqiao |
| contents | Anomaly analysis in surveillance videos is a crucial topic in computer vision. In recent years, multimodal large language models (MLLMs) have outperformed task-specific models in various domains. Although MLLMs are particularly versatile, their abilities to understand anomalous concepts and details are insufficiently studied because of the outdated benchmarks of this field not providing MLLM-style QAs and efficient algorithms to assess the model's open-ended text responses. To fill this gap, we propose a benchmark for crime surveillance video analysis with large models denoted as UCVL, including 1,829 videos and reorganized annotations from the UCF-Crime and UCF-Crime Annotation datasets. We design six types of questions and generate diverse QA pairs. Then we develop detailed instructions and use OpenAI's GPT-4o for accurate assessment. We benchmark eight prevailing MLLMs ranging from 0.5B to 40B parameters, and the results demonstrate the reliability of this bench. Moreover, we finetune LLaVA-OneVision on UCVL's training set. The improvement validates our data's high quality for video anomaly analysis. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_09325 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | A Benchmark for Crime Surveillance Video Analysis with Large Models Chen, Haoran Yi, Dong Cao, Moyan Huang, Chensen Zhu, Guibo Wang, Jinqiao Computer Vision and Pattern Recognition Anomaly analysis in surveillance videos is a crucial topic in computer vision. In recent years, multimodal large language models (MLLMs) have outperformed task-specific models in various domains. Although MLLMs are particularly versatile, their abilities to understand anomalous concepts and details are insufficiently studied because of the outdated benchmarks of this field not providing MLLM-style QAs and efficient algorithms to assess the model's open-ended text responses. To fill this gap, we propose a benchmark for crime surveillance video analysis with large models denoted as UCVL, including 1,829 videos and reorganized annotations from the UCF-Crime and UCF-Crime Annotation datasets. We design six types of questions and generate diverse QA pairs. Then we develop detailed instructions and use OpenAI's GPT-4o for accurate assessment. We benchmark eight prevailing MLLMs ranging from 0.5B to 40B parameters, and the results demonstrate the reliability of this bench. Moreover, we finetune LLaVA-OneVision on UCVL's training set. The improvement validates our data's high quality for video anomaly analysis. |
| title | A Benchmark for Crime Surveillance Video Analysis with Large Models |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2502.09325 |