AccesAI is an AI-driven tool designed to enhance web accessibility, ensuring a more inclusive digital experience for individuals with disabilities. Our solution focuses on the detection and remediation of common accessibility issues, adhering to WCAG standards.
- Alt Text Detection and Generation: Automatically identifies images without descriptive alt texts and generates meaningful replacements.
- Keyboard Navigation Checker: Ensures all interactive elements are accessible via keyboard navigation.
- Contrast Analysis: Detects poor color contrast and suggests WCAG-compliant adjustments.
- Unclear Labels Detection: Identifies and enhances unclear or missing labels for form and interactive elements.
- ARIA Attributes Integration: Validates and improves ARIA roles and properties for better assistive technology support.
Frontend: React
Backend: Flask
Languages: Python, JavaScript
Modules/Tools: Hugging Face, BeautifulSoup, Playwright, OpenCV
- Web Scraping and Parsing:
- Tools: BeautifulSoup, Playwright
- Extracted and evaluated HTML/CSS structures.
- Feature Implementation:
- Hugging Face Model: Analyzed images and detected missing alt texts.
- Text Analysis: Assessed labels and content clarity.
- Contrast Ratio Calculations: Ensured WCAG compliance.
- Keyboard Navigation Analysis: Verified tab order and focus-ability, and improved the same.
- ARIA Checker: Detected missing or misused ARIA attributes.
- Validation:
- Utilized OpenCV for screenshot comparisons.
- Input the website URL into the AccesAI interface.
- The system scrapes and analyzes the website’s structure and content.
- Accessibility issues are identified, categorized, and displayed in a user-friendly dashboard.
- Suggestions and automated fixes are applied, generating improved code.
- Visual validation ensures the changes enhance accessibility without compromising design.
Below is the high-level design diagram depicting the flow:

Follow these steps to set up and run AccesAI:
- Install necessary packages:
pip install uv uv add playwright uv run playwright install
- Navigate to the scripts folder:
cd scripts uv sync
- To run the server:
uv run app.py
- To run the frontend, refer to the frontend folder in the repository.
For a detailed understanding, please refer to the PowerPoint presentation linked below:
Download AccesAI PPT