Anthropometry and Pose Observation using Low-dimensional Latent Optimization (APOLLO)

The Anthropometry and Pose Observation using Low-dimensional Latent Optimization (APOLLO) system is a real-time, trauma-aware 3D mesh reconstruction and human pose estimation technology designed to support autonomous and semiautonomous medical systems in combat and disaster response environments. APOLLO enables robust inference of human body shape, pose, and mass from partial and multimodal sensor observations—including electro-optical (EO), depth, longwave infrared (LWIR), and near-infrared (NIR) imagery—even when the subject is occluded, injured, or viewed at close range. Unlike traditional models that assume intact anatomy and full-body visibility, APOLLO dynamically refines its mesh based on live input, allowing robotic platforms and medics to safely assess, manipulate, and treat casualties in real time.A key innovation of APOLLO is its ability to model and adapt to traumatic injuries—such as fractures, dislocations, hyperextensions, and amputations—allowing autonomous systems to generate actionable, anatomically consistent representations of wounded personnel. APOLLO also estimates the mass of limbs and body segments, supporting downstream robotic manipulation, casualty extraction, and context-aware medical decision-making such as drug dosing or injury prioritization. Unlike existing systems that assume ideal, intact geometry, APOLLO provides a robust and injury-informed foundation for real-time perception, enabling medics and autonomous platforms to safely engage with casualties in dynamic, high-risk environments.