Techniques for aligning visual content to three-dimensional signals

A non-transitory computer-readable medium having stored thereon instructions for causing a processing circuitry to execute a process, the process comprising:

• identifying a plurality of subjects in visual content, wherein the visual content is captured by a camera deployed at a site, wherein a position sensor is deployed with a target subject among the plurality of subjects, wherein the position sensor captures a plurality of signals representing a plurality of positions of the target subject at the site;
• transforming a plurality of points representing the plurality of positions of the target subject from a first feature space to a second feature space, wherein the first feature space is a feature space of the plurality of signals, wherein the second feature space is a feature space of the visual content, wherein transforming the plurality of points includes using an affine transformation, wherein the plurality of points representing the plurality of positions of the target subject in the first feature space form a curve, wherein transforming the plurality of points further includes classifying a plurality of portions of the curve into a plurality of classifications and mapping the plurality of portions of the curve to the plane of the second feature space based on a parameterization of the curve, wherein the plurality of points is transformed based on the plurality of classifications and the mapping;
• and
• mapping the plurality of positions of the target subject represented by the plurality of signals to a plurality of positions within the visual content based on the plurality of points in the second feature space.

A system for visual content alignment, comprising:

• a processing circuitry;
• and
• a memory, the memory containing instructions that, when executed by the processing circuitry, configure the system to:
• identify a plurality of subjects in visual content, wherein the visual content is captured by a camera deployed at a site, wherein a position sensor is deployed with a target subject among the plurality of subjects, wherein the position sensor captures a plurality of signals representing a plurality of positions of the target subject at the site;
• transform a plurality of points representing the plurality of positions of the target subject from a first feature space to a second feature space, wherein the first feature space is a feature space of the plurality of signals, wherein the second feature space is a feature space of the visual content, wherein transforming the plurality of points includes using an affine transformation, wherein the plurality of points representing the plurality of positions of the target subject in the first feature space form a curve, wherein transforming the plurality of points further includes classifying a plurality of portions of the curve into a plurality of classifications and mapping the plurality of portions of the curve to the plane of the second feature space based on a parameterization of the curve, wherein the plurality of points is transformed based on the plurality of classifications and the mapping;
• and
• map the plurality of positions of the target subject represented by the plurality of signals to a plurality of positions within the visual content based on the plurality of points in the second feature space.