Body Model Update from 2D Face Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current three-dimensional modeling of the human body requires expensive and large sensors, such as stereo imaging elements and depth sensing devices, which are not readily available in standard portable devices like smartphones and tablets.
Innovation Solution
A method to generate a dimensionally accurate body model using two-dimensional images from standard cameras, such as those found in smartphones, by processing segmented silhouettes and utilizing convolutional neural networks (CNNs) to predict body features and update the model with face-based features from 2D images, allowing for the creation of a full body model from partial images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If stereo imaging elements or depth sensing devices are used for three-dimensional body modeling, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent uses two-dimensional images as copies or proxies to represent three-dimensional body information. Instead of directly capturing 3D data with complex sensors, the system captures 2D images and uses machine learning models to infer 3D body characteristics, effectively replacing expensive depth sensing devices with standard 2D cameras.
Solution Approach 2:
The patent replaces the mechanical/optical system of stereo imaging and depth sensing with a computational system based on 2D image processing and machine learning. The convolutional neural networks and body composition models substitute for the physical depth-sensing hardware, achieving 3D body modeling through software-based inference rather than hardware-based measurement.
2Measurement precision
If stereo imaging elements or depth sensing devices are used for three-dimensional body modeling, then measurement precision is improved, but cost increases
Solution Approach 1:
The patent employs inexpensive 2D camera sensors that are already present in most smartphones and portable devices, replacing expensive dedicated depth-sensing hardware. The system uses readily available, low-cost imaging components combined with software processing to achieve accurate body modeling without requiring costly specialized equipment.
Solution Approach 2:
The patent makes standard 2D camera sensors perform the function of expensive 3D depth-sensing devices. By using machine learning models that can infer three-dimensional body characteristics from two-dimensional images, the system enables universal 3D body modeling capability across all devices with standard cameras, eliminating the need for specialized expensive sensors.
3Device complexity
If two-dimensional images are used to generate body models, then device complexity is reduced, but measurement precision deteriorates
Solution Approach 1:
The patent transforms the problem from direct 3D measurement to 2D image parameter analysis. By changing the approach from capturing 3D spatial data directly to analyzing 2D image parameters (pixel intensities, gradients, textures) and inferring 3D characteristics through learned relationships, the system achieves accurate body modeling with simpler 2D imaging equipment.
Solution Approach 2:
The patent introduces machine learning models (convolutional neural networks and body composition models) as intermediaries between the 2D image data and the 3D body model output. These intermediary computational layers process the 2D image information and translate it into accurate 3D body characteristics, bridging the gap between simple 2D input and precise 3D output.
Data Source
AI summary
Described are systems and methods directed to generation and subsequent update of a dimensionally accurate body model of a body, such as a human body, based on two-dimensional (“2D”) images of at least a portion of that body and/or face images of a face of the body. A user may use a 2D camera, such as a digital camera typically included in many of today's portable devices (e.g., cell phones, tablets, laptops, etc.) to produce body images that are used to generate a body model of the body of the user. Subsequently, the body model may be updated based on a face image of the face of the user, without requiring the user to provide another body image.