Dimensionality compression and expansion in deep neural networks
Datasets such as images, text, or movies are embedded in high-dimensional spaces.
However, in important cases such as images of objects, the statistical structure in the data …
However, in important cases such as images of objects, the statistical structure in the data …
Exploring Learned Representations of Neural Networks with Principal Component Analysis
Understanding feature representation for deep neural networks (DNNs) remains an open
question within the general field of explainable AI. We use principal component analysis …
question within the general field of explainable AI. We use principal component analysis …
Semi-self-supervised segmentation of oranges with small sample sizes
K Volle, P Ganesh, TF Burks… - 2020 ASABE Annual …, 2020 - elibrary.asabe.org
Visually detecting and masking fruit can be regarded as a vital first step in autonomous
harvesting. Given the constrained domain, general purpose segmentation networks are …
harvesting. Given the constrained domain, general purpose segmentation networks are …