Face Image Datasets
Face image datasets are crucial for training and evaluating facial recognition and analysis systems, driving research focused on improving accuracy, fairness, and privacy. Current research emphasizes mitigating biases stemming from demographic imbalances and data quality issues within these datasets, often employing techniques like knowledge distillation, synthetic data generation, and contrastive learning with models such as StyleGAN and Vision Transformers. These advancements aim to create more robust and equitable algorithms, impacting applications ranging from security and healthcare to forensic investigations and digital forensics.
Papers
November 12, 2022
September 27, 2022
July 27, 2022
May 23, 2022
May 3, 2022
January 9, 2022