17821620. THREE-DIMENSIONAL POSE DETECTION BASED ON TWO-DIMENSIONAL SIGNATURE MATCHING simplified abstract (QUALCOMM Incorporated)
Contents
THREE-DIMENSIONAL POSE DETECTION BASED ON TWO-DIMENSIONAL SIGNATURE MATCHING
Organization Name
Inventor(s)
Gary Franklin Gimenez of Bordeaux 33 (FR)
THREE-DIMENSIONAL POSE DETECTION BASED ON TWO-DIMENSIONAL SIGNATURE MATCHING - A simplified explanation of the abstract
This abstract first appeared for US patent application 17821620 titled 'THREE-DIMENSIONAL POSE DETECTION BASED ON TWO-DIMENSIONAL SIGNATURE MATCHING
Simplified Explanation
The patent application describes techniques for determining the pose of a three-dimensional deformable object using machine learning models and two-dimensional signatures.
- Inputs are provided to a machine learning model based on a computer-generated three-dimensional deformable object with a known pose.
- The machine learning model outputs a two-dimensional signature of the object based on the inputs.
- The two-dimensional signature is associated with the known pose of the computer-generated object.
- The pose of an actual three-dimensional deformable object is determined based on an image of the object and the associated two-dimensional signature.
- Potential Applications
- Augmented reality
- Virtual reality
- Robotics
- Animation
- Problems Solved
- Accurately determining the pose of deformable objects
- Improving object recognition and tracking
- Benefits
- Enhanced realism in virtual environments
- Improved object manipulation in robotics
- Streamlined animation processes
Original Abstract Submitted
Certain aspects of the present disclosure provide techniques for determining a pose of a three-dimensional deformable object. Embodiments include providing one or more inputs to a machine learning model based on a computer-generated three-dimensional deformable object that has a known pose. Embodiments include determining, based on one or more outputs from the machine learning model in response to the one or more inputs, a two-dimensional signature of the computer-generated three-dimensional deformable object. Embodiments include associating the two-dimensional signature with the known pose of the computer-generated three-dimensional deformable object. Embodiments include determining a respective pose of an actual three-dimensional deformable object based on an image of the actual three-dimensional deformable object and the associating.