Abstract: Triplet loss traditionally relies only on class labels and does not use all available information in multi-task scenarios where multiple types of annotations are available. This paper introduces a Multi-Annotation Triplet Loss (MATL) framework that extends triplet loss by… Read More
Tag: computer vision
Automated potato tuber mass estimation and grading with multiangle 2D images
Abstract: Estimating potato tuber mass and size grading with computer vision can help breeders, farmers, and potato processing units reduce manual labor for potato post-harvest handling through optimized technology. The objective of the study was to estimate potato tuber mass… Read More
Addressing the Inevitable Imprecision: Multiple Instance Learning for Hyperspectral Image Analysis
Abstract: In many remote sensing and hyperspectral image analysis applications, precise ground truth information is unavailable or impossible to obtain. Imprecision in ground truth often results from highly mixed or sub-pixel spectral responses over classes of interest, a mismatch between the precision… Read More