Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This study investigates the effectiveness of various segmentation architectures, including convolutional networks and vision transformers, on the B.O.V.I.D. dataset, which consists of high-resolution images of bovid dentition with imperfect segmentation masks. The research highlights the challenges posed by misaligned labels and evaluates preprocessing techniques aimed at improving model performance despite these imperfections. Key findings reveal that while quantitative metrics like Dice score and mIoU show limited improvement, the qualitative enhancement of predicted masks is significant, underscoring the importance of preprocessing in semantic segmentation tasks.
Imperfect annotations can drastically alter the qualitative output of segmentation models, even if quantitative metrics remain unchanged.
Semantic segmentation decomposes an image into distinct mask regions corresponding to different object categories, such as people, cars, signs or buildings. Advances in machine learning (ML) have shifted this task away from traditional rule-based heuristics such as edge detection, towards deep neural networks (DNN) that learn to classify pixels directly. However, semantic segmentation DNNs crucially depend on expertly designed mask targets to learn from, and imperfect or misaligned masks can interfere with a model's ability to learn effectively. 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. We evaluate a range of preprocessing and alignment techniques to mitigate the resulting label imperfections. We find that while these preprocessing choices have limited effect on quantitative metrics such as Dice score and mIoU, their qualitative impact on predicted masks is substantial.