Self-Normalized Density Map (SNDM) for Counting Microbiological Objects
This work addresses counting accuracy in microbiological imaging, which is important for researchers in biology and medicine, but it is incremental as it builds on existing density map methods.
The authors tackled the problem of counting microbiological objects in images by analyzing the statistical uncertainties of density map predictions, leading to the development of a self-normalized density map (SNDM) that corrects its output to accurately predict object counts, with SNDM outperforming the original model and achieving efficiency comparable to detector-based models like Faster and Cascade R-CNN.
The statistical properties of the density map (DM) approach to counting microbiological objects on images are studied in detail. The DM is given by U$^2$-Net. Two statistical methods for deep neural networks are utilized: the bootstrap and the Monte Carlo (MC) dropout. The detailed analysis of the uncertainties for the DM predictions leads to a deeper understanding of the DM model's deficiencies. Based on our investigation, we propose a self-normalization module in the network. The improved network model, called \textit{Self-Normalized Density Map} (SNDM), can correct its output density map by itself to accurately predict the total number of objects in the image. The SNDM architecture outperforms the original model. Moreover, both statistical frameworks -- bootstrap and MC dropout -- have consistent statistical results for SNDM, which were not observed in the original model. The SNDM efficiency is comparable with the detector-base models, such as Faster and Cascade R-CNN detectors.