One of the most important tasks for ensuring safe autonomous driving systems is accurately detecting road traffic lights and accurately determining how they impact the driver's actions. In various real-world driving situations, a scene may have numerous traffic lights with varying levels of relevance to the driver, and thus, distinguishing and detecting the lights that are relevant to the driver and influence the driver's actions is a critical safety task. This paper proposes a traffic light detection model which focuses on this task by first defining salient lights as the lights that affect the driver's future decisions. We then use this salience property to construct the LAVA Salient Lights Dataset, the first US traffic light dataset with an annotated salience property. Subsequently, we train a Deformable DETR object detection transformer model using Salience-Sensitive Focal Loss to emphasize stronger performance on salient traffic lights, showing that a model trained with this loss function has stronger recall than one trained without.
By means of a laborious technique, based on the use of a projection microscope and a sensitive electric photometer, it has been shown that both the areas and transparencies of stained red blood cells may be determined. From the few observations completed, it appears that, statistically, these attributes are positively correlated in normal human blood and their regression line is straight. In pernicious anemia the A/T correlation is negative. Cases of secondary anemia may show either positive or negative A/T correlations. The pathological blood specimens studied all showed non-linear A/T regression lines.
This is a continuation of observations published in this Journal (4). It deals with an electrical photometer of high sensitivity but unaffected by red light. This is used in conjunction with the projection microscope and applied to a study of red blood cells stained with basic fuchsine. The measurements reported are a complex of staining intensity and area. These are compared with area estimations alone and it is shown that the former constitutes a greater variable than the latter. This is particularly noticeable when dealing with anemic blood. The relationship between primary and secondary anemia of man is touched upon.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.