AI Body Measurement Estimation for Virtual Try-On

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing virtual environments struggle to accurately simulate real-world scenarios, particularly in recreating human body measurements and garment fit, due to limitations in specialized hardware and incomplete data input from users.

Innovation Solution

An artificial intelligence model trained on human body measurement data and garment specifications is used to estimate body measurements and determine garment sizes based on minimal user input, without the need for specialized hardware.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If specialized hardware and extensive software-based user configuration are used to create virtual environments, then realism and immersion are improved, but device complexity and cost increase

Engineering Contradiction:
ImproverealismVSAvoidhardware complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent creates a virtual avatar that copies and represents the user's physical body characteristics, including measurements and appearance, without requiring specialized hardware. The avatar replicates real-world body data through computational models trained on user-provided information, achieving realism through digital representation rather than complex hardware simulations.

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent replaces mechanical measurement systems with artificial intelligence models. Instead of using specialized hardware to physically measure body dimensions, the system uses AI algorithms trained on body measurement data to estimate measurements from user inputs, substituting mechanical complexity with computational intelligence.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If AI models are trained on comprehensive body measurement data, then measurement precision is improved, but loss of information about user input requirements increases

Engineering Contradiction:
Improvebody measurement accuracyVSAvoiduser input information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent applies partial action by requiring users to provide only specific body attributes and measurements relevant to the task, rather than demanding complete body scans or extensive data collection. The AI model processes these partial inputs to generate comprehensive body measurements, balancing data requirements with user convenience and information preservation.

Inventive Principle:
Principle #16Partial or excessive action

3Measurement precision

If virtual environments use extensive software-based user configuration, then measurement precision is improved, but ease of operation deteriorates

Engineering Contradiction:
Improvebody measurement accuracyVSAvoiduser configuration ease
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The patent implements self-service by enabling users to input their own body attributes and measurements directly into the system. The AI model automatically processes this user-provided information to generate comprehensive body measurements and avatar representations, eliminating the need for complex external measurement tools or configurations while maintaining accuracy.

Inventive Principle:
Principle #25Self-service

Data Source

PatentUS20250157166A1Artificial intelligence systems for generation of human body part measurements and human body fit information
Publication Date: 2025.05.15 BOLD METRICS INC
  • US20250157166A1 patent drawing
  • US20250157166A1 patent drawing
  • US20250157166A1 patent drawing

AI summary

Systems and methods are disclosed for generation of human body part measurements and human body fit information. In some embodiments, a body-realistic human user avatar is created based on generated body measurements for the user. The disclosure includes use of artificial intelligence models to generate extensive body measurements for the user based on minimal physical measurements provided by the user as a training input to the artificial intelligence model. Aspects of the present disclosure address shortcomings present in existing virtual environment approaches that fail to accurately represent real-world human body characteristics or require extensive user-facing hardware components.