Update README.md

README.md

@@ -58,31 +58,94 @@ configs:
     path: data/test_primary_oe-*
 ---
 
-# Towards Video Turing Test (Video-TT): Video Comprehension and Reasoning Benchmark with Complex Visual Narratives
+# Towards Video Turing Test (Video-TT): A Benchmark for Video Comprehension and Reasoning with Complex Visual Narratives
 
-Video-TT comprises 1,000 YouTube Shorts videos, each with one open-ended question and four adversarial questions that probe visual and narrative complexity.
+Video-TT comprises 1,000 YouTube videos, each paired with one open-ended question and four adversarial questions designed to probe visual and narrative complexity.
 
 ## 🚀 What's New
-- **\[2025.03\]** We release the benchmark!
+- **[2025.03]** We release the benchmark!
 
-## 1. Why we need a new benchmark like Video-TT?
+## 1. Why Do We Need a New Benchmark Like Video-TT?
 
-- **Sampling v.s Understanding:** Current video understanding benchmark do not clearly distinguish between errors caused by insufficient frame sampling and errors caused by failures in actual video understanding.  We ensure that each question in Video-TT can be well answered with uniformly sampled 80 frames where most current video model can easily handle.v
-- **Pursing human level video understanding:** We carefully select Q&A, where human can achieve 87.5 points, but even the best model can only achieve 47.5. In comparison, without sampling issue, current video model can achieve more than 85 points on video understanding benchmark [1]
+- **Sampling vs. Understanding:** Current video understanding benchmarks do not clearly distinguish between errors caused by insufficient frame sampling and errors due to failures in actual video comprehension. We ensure that each question in Video-TT can be answered with 80 uniformly sampled frames—frames that most current video models can easily process.
+- **Pursuing Human-Level Video Understanding:** We carefully select Q&A pairs where humans achieve an 87.5% accuracy rate, while the best models only reach 47.5%. In contrast, when sampling is not a limiting factor, current video models score above 85% on existing video understanding benchmarks [1].
 
-## 2. How to ensure the quality of Video-TT?
+## 2. Dataset Summary
 
+We introduce five question prototypes. 
 
-## 3. How to evaluate on Video-TT
+### Dataset Statistics
+- **Number of videos:** 1,000, primarily in formats like YouTube Shorts.
+- **Number of Q&A pairs:** Each video is paired with five questions that probe different aspects of the same content, designed to challenge model robustness in adversarial scenarios:
+  - Primary open-ended question
+  - Paraphrased open-ended question
+  - Correctly-led open-ended question
+  - Wrongly-led open-ended question
+  - Multiple-choice question
 
-## 4. Dataset summary
+One example of five question types is shown below:
+<img src="dataset_expansion_page-0001.jpg" style="width:30%;">
 
+### Evaluation Metrics
 
+#### Correctness:
 
-[1] Fu, Chaoyou, et al. "Video-mme: The first-ever comprehensive evaluation benchmark of multi-modal llms in video analysis." arXiv preprint arXiv:2405.21075 (2024).
+Measures accuracy for each question type. We use Qwen2.5-72B as the judge for open-ended questions and a rule-based method for multiple-choice questions.
 
+#### Robustness:
 
+Let:
+- A_primary_correct be the set of videos where the primary open-ended question is answered correctly.
+- A_paraphrased_correct be the set of videos where the paraphrased open-ended question is answered correctly.
+- A_correctly_led_correct be the set of videos where the correctly-led open-ended question is answered correctly.
+- A_wrongly_led_correct be the set of videos where the wrongly-led open-ended question is answered correctly.
+- A_multiple_choice_correct be the set of videos where the multiple-choice question is answered correctly.
 
+The set of videos where all five questions are answered correctly, denoted as A_full_correct, is the intersection of all these sets:
 
+A_full_correct = A_primary_correct ∩ A_paraphrased_correct ∩ A_correctly_led_correct ∩ A_wrongly_led_correct ∩ A_multiple_choice_correct
 
+Thus, the **Robustness Score (RB)** becomes:
 
+R = |A_full_correct| / |A_primary_correct|
+
+Where |A| denotes the size of the set A, representing the number of videos in that set.
+
+
+## 3. Ensuring the Quality of Video-TT
+
+<img src="annotation_pipeline_page-0001.jpg" style="width:50%;">
+
+The Video-TT annotation process consists of four stages:
+
+- **Stage 1:** We select complex videos that allow for challenging questions. (Refer to the paper for the definition of complexity factors.)
+- **Stage 2:** We prompt several vision-language models (VLMs) to check whether a question can be answered. If a model answers it correctly, we remove the question.
+- **Stage 3:** For each remaining question, we provide an answer and an explanation.
+- **Stage 4:** We manually verify that each question can be answered using 80 uniformly sampled frames.
+
+For more details, please refer to the paper.
+
+## 4. How to Evaluate on Video-TT
+
+(To be detailed in the paper.)
+
+## 5. Leaderboard
+
+For the latest leaderboard, please refer to **XXX**. You can submit your evaluations there. Some results are produced by us, while others are submitted by external researchers.
+
+To reproduce our results, please check **XXX** for evaluation scripts.
+
+<img src="771742752771_.pic.jpg" style="width:50%;">
+
+## 6. Dataset Maintenance
+
+If you find any mistakes in the dataset, please submit the corresponding `question_id` to our issue page. Our team is committed to maintaining this dataset in the long run to ensure its quality.
+
+## 7. Acknowledgments
+
+(Include acknowledgments as needed.)
+
+
+
+
+[1] Fu, Chaoyou, et al. "Video-mme: The first-ever comprehensive evaluation benchmark of multi-modal llms in video analysis." arXiv preprint arXiv:2405.21075 (2024).