Segmentation of Bovid Dentition Under Imperfect Annotations: A Comparative Study of Convolutional and Attention Models
This paper presents a comparative study of segmentation architectures, ranging from convolutional backbones to vision transformers, applied to the B.O.V.I.D. dataset, a corpus of high-resolution bovid dental photographs paired with hand-made segmentation masks not originally designed for ML-based training, and evaluates a range of preprocessing and alignment techniques to mitigate the resulting label imperfections.
Keith G. Mills, Evan B. Sanders, Gregory J. Matthews et al.
· 0 citations