Transferring Landmark Annotations for Cross-Dataset Face Alignment
This addresses dataset bias for face alignment researchers, offering a solution to combine existing datasets without re-labeling, though it is incremental as it builds on known issues in object recognition.
The paper tackles the problem of dataset bias in face alignment, where training on one database and testing on another leads to poor performance, and proposes a method to bridge annotation spaces between databases, enabling dataset fusion and showing improved cross-dataset alignment on databases like LFW, AFLW, LFPW, and HELEN.
Dataset bias is a well known problem in object recognition domain. This issue, nonetheless, is rarely explored in face alignment research. In this study, we show that dataset plays an integral part of face alignment performance. Specifically, owing to face alignment dataset bias, training on one database and testing on another or unseen domain would lead to poor performance. Creating an unbiased dataset through combining various existing databases, however, is non-trivial as one has to exhaustively re-label the landmarks for standardisation. In this work, we propose a simple and yet effective method to bridge the disparate annotation spaces between databases, making datasets fusion possible. We show extensive results on combining various popular databases (LFW, AFLW, LFPW, HELEN) for improved cross-dataset and unseen data alignment.