Overview
HowDiverse.is - use AI to analyse diversity in organisations
HowDiverse uses image recognition and AI to analyse gender and ethnic diversity in organisations. The HowDiverse Chrome extension extracts faces from photos of employees on websites such as About Us, and then runs them through a Machine Learning (Deep Learning) pipeline. This pipeline detects all the faces in the images, even multiple faces in a single image, and extracts them. The extracted faces are then analysed for gender and ethnicity using another pipeline. How accurate is it? The program is still in beta (so we would really value feedback) but the model behind it has an accuracy of 97% which is better than most humans. We're adding ability for users to correct wrong results, and we will use these corrections to continuously improve the model. Why this tool? The tool helps ensure transparency and accountability for organisations when it comes to diversity, equity, and inclusion (DEI). What are the thresholds? The threshold for not being flagged red for Black is 4%, which is based on the percentage of black population in the UK. The threshold for gender is - obviously - 50% women. Why only Black? HowDiverse focuses on Black folks due to their consistently low representation, but other ethnicities breakdown is shown on our website https://howdiverse.is. Why only binary gender and ethnicity? HowDiverse plans to develop a platform where companies can gather self-identification data beyond binary gender and ethnicity. Contact info@nyala.dev or join the waiting list at https://howDiverse.is if interested.
Details
- Version0.0.13
- UpdatedOctober 12, 2023
- Size423KiB
- LanguagesEnglish
- Developer
- TraderThis developer has identified itself as a trader per the definition from the European Union.
Privacy
This developer declares that your data is
- Not being sold to third parties, outside of the approved use cases
- Not being used or transferred for purposes that are unrelated to the item's core functionality
- Not being used or transferred to determine creditworthiness or for lending purposes
