Awesome
EvalBoard 🎥📊
EvalBoard is a web application built using Gradio that displays a gallery of videos generated by various methods. It allows users to navigate through the gallery, view videos, and read associated captions.
Installation
To run EvalBoard locally, follow these steps:
-
Clone the repository:
git clone https://github.com/evalcrafter/EvalBoard
-
Navigate to the project directory:
cd EvalBoard
-
Install the required dependencies:
pip install -r requirements.txt
Data Preparation
Before running the EvalBoard application, you need to prepare the video data by following these steps:
-
Create a directory named
static
in the project root if it doesn't already exist. -
Inside the
static
directory, create subdirectories for each category of videos. The category names are specified in theshowcase
function in theEvalBoard.py
file. You can also replace the current names with yours. -
Place the video files in their respective category subdirectories. The videos should be named using a four-digit index followed by the
.mp4
extension. For example, the first video for the categoryVideoCrafter2
should be named0000.mp4
, the second video should be named0001.mp4
, and so on. Default to have 700 videos. -
Place the prompts files corresponding to video names in
prompts
, using the same name with.txt
extension. You may delete the current files first.
Usage
To start EvalBoard, execute the following command within the project directory:
python app.py
Once the application is running, open your web browser and visit http://localhost:8000
to access EvalBoard.
Upon launching EvalBoard, you will see a gallery of videos with their associated captions. The application provides several ways to navigate through the gallery:
- Beginning: Clicking the "Beginning" button takes you to the first page.
- Previous: Clicking the "Previous" button takes you to the previous page.
- Next: Clicking the "Next" button takes you to the next page.
- End: Clicking the "End" button takes you to the last page.
- Slider: You can use the slider to directly navigate to a specific page.
As you navigate through the gallery, the videos and captions will dynamically update to reflect the current page. Simply click on a video to play it within the application.
Citation
If you find this repository helpful, please consider citing it in your research:
@article{liu2023evalcrafter,
title={Evalcrafter: Benchmarking and evaluating large video generation models},
author={Liu, Yaofang and Cun, Xiaodong and Liu, Xuebo and Wang, Xintao and Zhang, Yong and Chen, Haoxin and Liu, Yang and Zeng, Tieyong and Chan, Raymond and Shan, Ying},
journal={arXiv preprint arXiv:2310.11440},
year={2023}
}