K-Origins: Better Colour Quantification for Neural Networks
This work addresses a domain-specific problem in computer vision for image segmentation tasks, offering an incremental improvement in network performance.
The paper tackles the problem of improving semantic segmentation accuracy in neural networks for scenarios involving low signal-to-noise ratios or objects identical in shape but varying in color, by introducing K-Origins, a layer that enhances color quantification, resulting in demonstrated improvements over 250 encoder-decoder networks trained on 16-bit synthetic data.
K-Origins is a neural network layer designed to improve image-based network performances when learning colour, or intensities, is beneficial. Over 250 encoder-decoder convolutional networks are trained and tested on 16-bit synthetic data, demonstrating that K-Origins improves semantic segmentation accuracy in two scenarios: object detection with low signal-to-noise ratios, and segmenting multiple objects that are identical in shape but vary in colour. K-Origins generates output features from the input features, $\textbf{X}$, by the equation $\textbf{Y}_k = \textbf{X}-\textbf{J}\cdot w_k$ for each trainable parameter $w_k$, where $\textbf{J}$ is a matrix of ones. Additionally, networks with varying receptive fields were trained to determine optimal network depths based on the dimensions of target classes, suggesting that receptive field lengths should exceed object sizes. By ensuring a sufficient receptive field length and incorporating K-Origins, we can achieve better semantic network performance.