Double Trouble? Impact and Detection of Duplicates in Face Image Datasets
This addresses data quality issues for researchers in facial biometrics, but is incremental as it applies existing techniques to new datasets.
The study tackled the problem of duplicate face images in web-scraped datasets by developing a detection method using hashes and preprocessing, finding hundreds to hundreds of thousands of duplicates in four out of five datasets, with minor impact on recognition and quality assessment results after removal.
Various face image datasets intended for facial biometrics research were created via web-scraping, i.e. the collection of images publicly available on the internet. This work presents an approach to detect both exactly and nearly identical face image duplicates, using file and image hashes. The approach is extended through the use of face image preprocessing. Additional steps based on face recognition and face image quality assessment models reduce false positives, and facilitate the deduplication of the face images both for intra- and inter-subject duplicate sets. The presented approach is applied to five datasets, namely LFW, TinyFace, Adience, CASIA-WebFace, and C-MS-Celeb (a cleaned MS-Celeb-1M variant). Duplicates are detected within every dataset, with hundreds to hundreds of thousands of duplicates for all except LFW. Face recognition and quality assessment experiments indicate a minor impact on the results through the duplicate removal. The final deduplication data is publicly available.