Can the accuracy bias by facial hairstyle be reduced through balancing the training data?
Kagan ÖztürkHaiyu WuKevin W. BowyerPublished in: CoRR (2024)
Keyphrases
- training data
- human faces
- high accuracy
- classification accuracy
- error rate
- data sets
- prediction accuracy
- decision trees
- supervised learning
- landmark detection
- highly accurate
- facial images
- computational efficiency
- labeled data
- training set
- face detection
- computational cost
- face recognition
- learning algorithm
- training process
- facial expressions
- training examples
- support vector machine
- computer vision
- unseen data
- age estimation
- machine learning