Winner (MLH x ALL-In Hackathon - Best Accessibility Award sponsored by Fidelity in helping make tech more accessible)
The Accessible Data Visualization Toolkit is a Python-based project designed to create data visualizations with accessibility features. The project uses libraries such as Matplotlib, Seaborn, gTTS, Pydub to create a variety of data visualizations and make them accessible to individuals with visual impairments.
The main features of the project include:
Ability to load CSV files from a GitHub repository Creation of accessible data visualizations such as histograms, scatter plots, and line charts Ability to add alternative text descriptions to visual elements Adjustable font sizes and colors to suit user preferences Compatibility with screen readers through the use of gTTS and Pydub libraries Option to save the plots as PNG files with customizable output file names The project begins by asking the user for a GitHub link to load the CSV file. After loading the CSV file, the user is presented with a list of available columns and can select a column for the histogram. The user can also choose the color and title for the histogram and specify the font size they are comfortable with. The histogram is created using Matplotlib, and descriptive alt text is added to make it accessible. The user can also choose to create a scatter plot or line chart using Seaborn and specify the columns for the x-axis and y-axis. Descriptive alt text is added to make these visualizations accessible as well.
The project is designed with accessibility in mind, with adjustable font sizes and colors, and text-to-speech functionality to ensure compatibility with screen readers. Users can also save the plots as PNG files with customizable output file names.
The text-to-speech functionality included in this code can be helpful for visually impaired individuals to hear the instructions and feedback provided by the script during the data visualization process. This can potentially make it easier for them to navigate and understand the code.