Publications by authors named "Emily D Bello-Pardo"
Article Synopsis
- Image recognition systems can process images without needing expert input, but they often come with biases.
- The study evaluates gender biases in commercial image recognition by analyzing photos of U.S. Congress members and their Twitter images, revealing that women receive more appearance-related labels while being recognized less often than men.
- This biased output not only impacts the visibility of women but also reinforces harmful stereotypes, raising concerns about the reliability of data derived from these systems.
