Leaderboard for text-to-video generation models
Generate detailed data reports
Generate plots for GP and PFN posterior approximations
Browse and submit evaluation results for AI benchmarks
Analyze and visualize Hugging Face model download stats
Display a treemap of languages and datasets
Gather data from websites
Analyze and visualize car data
Make RAG evaluation dataset. 100% compatible to AutoRAG
Try the Hugging Face API through the playground
View and compare pass@k metrics for AI models
This is a timeline of all the available models released
Visualize dataset distributions with facets
VideoScore Leaderboard is a tool designed to compare and analyze the performance of text-to-video generation models. It provides a clear and organized way to display leaderboard tables, showcasing video scores and evaluation data. This tool is essential for researchers and developers to track progress, identify top-performing models, and make data-driven decisions.
• Score Display: Shows video scores in a structured leaderboard format.
• Model Comparison: Allows side-by-side comparison of different models.
• Real-Time Data Updates: Reflects the latest evaluation results.
• Interactive Tables: Enables sorting, filtering, and searching functionalities.
• Data Visualization: Includes charts to represent trends and performance metrics.
• Customization: Users can filter by specific criteria or metrics.
What is a "model" in the context of VideoScore Leaderboard?
A model refers to a specific text-to-video generation algorithm or system being evaluated.
Can I customize the leaderboard to show only specific models?
Yes, users can filter the leaderboard to display only the models they are interested in.
How are the video scores calculated?
Video scores are calculated using predefined evaluation metrics, which may vary depending on the dataset or use case.