Update app.py

app.py

@@ -7,12 +7,17 @@ BANNER = f'<div style="display: flex; justify-content: space-around;"><img src="
 INTRODUCTION_TEXT = """
 📖**Open Universal Arabic ASR Leaderboard**📖 benchmarks multi-dialect Arabic ASR models on various multi-dialect datasets.
 \nApart from the WER%/CER% for each test set, we also report the Average WER%/CER% and rank the models based on the Average WER, from lowest to highest.
-\nTo reproduce the benchmark numbers and request a model that is not listed, you can launch an issue/PR in our GitHub repo😊.
-\nFor more detailed analysis such as models' robustness, speaker adaption, model efficiency and memory usage, please check our paper.
+\nTo reproduce the benchmark numbers and request a model that is not listed, you can launch an issue/PR in our [GitHub repo](https://github.com/Natural-Language-Processing-Elm/open_universal_arabic_asr_leaderboard)😊.
+\nFor more detailed analysis such as models' robustness, speaker adaption, model efficiency and memory usage, please check our [paper](https://arxiv.org/pdf/2412.13788).
 """
 
 CITATION_BUTTON_TEXT = """
-???
+@article{wang2024open,
+  title={Open Universal Arabic ASR Leaderboard},
+  author={Wang, Yingzhi and Alhmoud, Anas and Alqurishi, Muhammad},
+  journal={arXiv preprint arXiv:2412.13788},
+  year={2024}
+}
 """
 
 METRICS_TAB_TEXT = METRICS_TAB_TEXT = """
@@ -20,7 +25,7 @@ METRICS_TAB_TEXT = METRICS_TAB_TEXT = """
 We report both the Word Error Rate (WER) and Character Error Rate (CER) metrics.
 ## Reproduction
 The Open Universal Arabic ASR Leaderboard will be a continuous benchmark project. 
-\nWe open-source the evaluation scripts at our GitHub repo.
+\nWe open-source the evaluation scripts at our [GitHub repo](https://github.com/Natural-Language-Processing-Elm/open_universal_arabic_asr_leaderboard).
 \nPlease launch a discussion in our GitHub repo to let us know if you want to learn about the performance of a new model.
 
 ## Benchmark datasets
@@ -33,8 +38,8 @@ The Open Universal Arabic ASR Leaderboard will be a continuous benchmark project
 | [MGB-2](http://www.mgb-challenge.org/MGB-2.html)                                                | Unspecified    | 9.6         |
 
 ## In-depth Analysis
-We also provide a comprehensive analysis of models' robustness, speaker adaptation, inference efficiency and memory consumption.
-\nPlease check our paper to learn more.
+We also provide a comprehensive analysis of the models' robustness, speaker adaptation, inference efficiency and memory consumption.
+\nPlease check our [paper](https://arxiv.org/pdf/2412.13788) to learn more.
 """