This is the official repository for Video-BLADE, a framework for efficient video generation through joint training of adaptive sparse attention and step distillation.
Visit our project homepage: http://ziplab.co/BLADE-Homepage/
- Academic-Focused Design: A modern and professional layout tailored for academic presentation.
- Responsive Layout: Perfectly adapts to desktops, tablets, and mobile devices.
- Interactive Features: Smooth scrolling navigation, image gallery, and video showcases.
- Performance Optimized: Fast loading times, lazy loading for images, and optimized animations.
- Accessibility: Supports keyboard navigation and is screen reader compatible.
homepage/
├── index.html # Main page
├── css/
│ ├── style.css # Main stylesheet
│ └── responsive.css # Responsive styles
├── js/
│ ├── main.js # Main JavaScript functionalities
│ └── gallery.js # Image gallery functionality
├── images/ # Image assets
│ ├── method_overview.png
│ ├── Table1.png
│ ├── Table2.png
│ └── ...
├── videos/ # Video assets
│ ├── cogvideox_baseline_*.mp4
│ ├── cogvideox_blade_*.mp4
│ └── ...
├── assets/ # Other assets
├── test.html # Test page
└── README.md # README document
cd homepage
python3 -m http.server 8000
Then, open your browser and navigate to: http://localhost:8000
Double-click the index.html
file. Note: some interactive features may be restricted by browser security policies when opening the file directly.
This project has been deployed to GitHub Pages and is accessible at the following URL: http://ziplab.co/BLADE-Homepage/
- HTML5: Semantic markup
- CSS3: Modern styling features (Grid, Flexbox, Variables)
- JavaScript: Vanilla ES6+
- Fonts: Inter (from Google Fonts)
- Icons: Font Awesome
- Lazy loading for images
- CSS and JavaScript minification (for production)
- Modern browser feature detection
- Progressive enhancement design
Video-BLADE Homepage - A professional, aesthetic, and efficient project page for academic papers.