AB Lab Projects

This project explores how slang is understood and recognized in voice-enabled technology by Black people. The goal of the project is to understand how slang is used in voice techhnology and how these tools can be more inclusive of slang terms.

This project explores empathy and microaggressions that humans experience in person-to-person speech. Microaggressions are brief, subtle, verbal, behavioral, or environmental indignities, which communicate negative prejudicial insults toward any group, particularly culturally marginalized groups. AI Models are rarely built containing multimodal examples that provide more context into abusive speech. We believe this research will help create smarter Diversity and Inclusion training tools.

This project is funded by the NSF, Amazon, and Salesforce.

This project explores the improvement and analysis of the state-of-the-art of biometric identification and recognition using ear images. The research will involve an end-to-end study of ear segmentation, ear similarity, and existing datasets. Currently, the work is focused on creating a novel, inclusive dataset and researching the state-of-the-art of ear similarity.

This project funded by the NSF, NSA, NGA and Northrop Grumman examines how codeswitching is done to hide sentiment of social media data