Our goal is to highlight once more the gender bias appearing naturally in Neural Networks, this time in the context of age recognition using face data. After demonstrating its impact and discussing its origin, we will attempt to develop methodologies to counter its influence on the output of deep convolutional neural networks.
We will be relying on the the OUI-Adience Face Image Project to train, validate and test our architecture.
- Review of Deep Learning Techniques for Gender Classification in Images
- Deep Learning for Face Recognition: Pride or Prejudiced?
- Determining Bias in Machine Translation with Deep Learning Techniques
- Gender Bias in Contextualized Word Embeddings
- Exploring Automatic Face Recognition on Match Performance and Gender Bias for Children
- SensitiveNets: Learning Agnostic Representations with Application to Face Recognition
- Wondering is enough: Uncertainty about category information undermines face recognition
- Robust Face Analysis Employing Machine Learning Techniques for Remote Heart Rate Estimation and towards Unbiased Attribute Analysis
- Recognizing human facial expressions with machine learning
Image credits : UC Berkeley