Add paper and GitHub links, update task categories

README.md

@@ -1,5 +1,40 @@
 ---
 license: cc-by-nc-sa-4.0
 task_categories:
+- video-text-to-text
 - question-answering
----
+---
+
+# CUVA: Causation Understanding of Video Anomaly
+
+This is the official dataset repository for the paper **"Uncovering What, Why and How: A Comprehensive Benchmark for Causation Understanding of Video Anomaly"** (CVPR 2024).
+
+[**Paper**](https://huggingface.co/papers/2405.00181) | [**GitHub Repository**](https://github.com/fesvhtr/CUVA)
+
+## Introduction
+Video Anomaly Understanding (VAU) aims to automatically comprehend unusual occurrences in videos. While existing VAU benchmarks primarily concentrate on anomaly detection and localization, CUVA focuses on more practical applications by addressing three crucial questions:
+- **What** anomaly occurred? (Anomaly type, start/end times, and event descriptions)
+- **Why** did it happen? (Natural language explanations for the cause of an anomaly)
+- **How** severe is it? (Free text reflecting the effect of the abnormality)
+
+The benchmark involves three sets of human annotations and introduces **MMEval**, a novel evaluation metric designed to better align with human preferences for causation understanding.
+
+## Dataset Structure
+The dataset consists of video anomaly instances with annotations stored in JSON format.
+As noted in the official repository, there are 4 zip files and 1 json file. Users should unzip the files and organize them into a `data` folder for use with the provided baseline and evaluation scripts.
+
+## Citation
+If you find this work useful for your research, please consider citing:
+
+```bibtex
+@INPROCEEDINGS{CUVA,
+  author={Du, Hang and Zhang, Sicheng and Xie, Binzhu and Nan, Guoshun and Zhang, Jiayang and Xu, Junrui and Liu, Hangyu and Leng, Sicong and Liu, Jiangming and Fan, Hehe and Huang, Dajiu and Feng, Jing and Chen, Linli and Zhang, Can and Li, Xuhuan and Zhang, Hao and Chen, Jianhang and Cui, Qimei and Tao, Xiaofeng},
+  booktitle={2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, 
+  title={Uncovering what, why and How: A Comprehensive Benchmark for Causation Understanding of Video Anomaly}, 
+  year={2024},
+  volume={},
+  number={},
+  pages={18793-18803},
+  keywords={Measurement;Annotations;Surveillance;Natural languages;Benchmark testing;Traffic control;Pattern recognition;Anomaly Video;Large Language Model},
+  doi={10.1109/CVPR52733.2024.01778}}
+```