Agile Sciences (United States)

Data labeling is often the most time consuming and expensive component of new automatic target recognition (ATR) model development in remote sensing. Synthetic data have long been desired for ATR model development in scenarios with limited real data available but have not been fully adopted in defense due to compute constraints and truth limitations within critical applications. This work combines AgileView’s synthetic data generation platform with Bedrock Research’s remote sensing foundation models to perform automated active learning. We seed initial model training on an example use case of maritime object classification solely with synthetic data, and then an iterative feedback loop of foundation model fine tuning with exclusively real data fully automates the active learning process and minimizes the human hours required for comprehensive data labeling. Furthermore, this technique introduces novel measures of label and image stability to automatically quantify the reliability of an auto-generated label. This provides a fast and trustworthy approach to ATR data and model curation for widespread defense applications.

Biofilm formation is important in the pathogenesis of many bacterial infections of both humans and animals. Coagulase-negative staphylococci (CNS) are significant biofilm-producing pathogens found in human foreign-body infections and ruminant mastitis. We have previously shown differences in biofilm production by representatives of various genotypes of Staphylococcus aureus causing bovine mastitis. We further demonstrated that 2-aminoimidazole compounds have anti-biofilm and antibacterial potentiation properties when tested in vitro using these isolates. The CNS are the bacteria most commonly isolated from goat milk, and they can be important pathogens in the goat mammary gland. Our purpose was to evaluate the biofilm-forming potential of representatives of genotypes of the most common CNS we found in goat mastitis. We hypothesized differences in biofilm-producing potential among goat CNS isolates and that biofilm formation may reflect pathogenic potential. Using the most prolific biofilm-forming CNS isolate, we hypothesized that 2-aminoimidazole compounds would inhibit biofilm production and potentiate antibacterial activity.

A polymeric composite containing a 2-aminoimidazole derivative was synthesized. It was found that this polymer was resistant to biofilm colonization by Acinetobacter baumannii, no leaching of the 2-aminoimidazole derivative was observed after 2 weeks of treatment with deionized water, and the resulting polymer was not hemolytic.