Corpus ID: 211096751

Machines Learn Appearance Bias in Face Recognition

@article{Steed2020MachinesLA,
  title={Machines Learn Appearance Bias in Face Recognition},
  author={Ryan Steed and Aylin Caliskan},
  journal={ArXiv},
  year={2020},
  volume={abs/2002.05636}
}
We seek to determine whether state-of-the-art, black box face recognition techniques can learn first-impression appearance bias from human annotations. With FaceNet, a popular face recognition architecture, we train a transfer learning model on human subjects' first impressions of personality traits in other faces. We measure the extent to which this appearance bias is embedded and benchmark learning performance for six different perceived traits. In particular, we find that our model is better… Expand
How (Not) to Measure Bias in Face Recognition Networks
Mitigating Biases in Multimodal Personality Assessment

References

SHOWING 1-10 OF 39 REFERENCES
FaceNet: A unified embedding for face recognition and clustering
The functional basis of face evaluation
First Impressions: A Survey on Vision-Based Apparent Personality Trait Analysis
MS-Celeb-1M: A Dataset and Benchmark for Large-Scale Face Recognition
Predictive Inequity in Object Detection
First Impressions
Design of an explainable machine learning challenge for video interviews
Facing faces: studies on the cognitive aspects of physiognomy.
  • R. Hassin, Y. Trope
  • Psychology, Medicine
  • Journal of personality and social psychology
  • 2000
...
1
2
3
4
...