End-to-end personalization and engagement platform
built for the new era
of experiential commerce
-
- Acknowledged in Gartner® Hype Cycle™ for Retail Technologies, 2021
-
- Awarded ‘Best In-Store Technology’ at the Digiday Technology Awards
-
- Winner of the LVMH Innovation Award, 2019
-
- RTIH Innovation Awards winner for ‘Retailer/Technology Supplier Relationship’
Proprietary AI-first technology enabling a unique virtual dressing room experience
Our unique virtual dressing room experience combines accurate size recommendations with photo-realistic virtual try-on, powered by data. Сapture measurements, process data, and deliver personalized results to each and every customer — all through one simple, streamlined solution.
Our platform superpowers
Mini-app technology
(App clips and Instant apps)
Lightweight and fast way for the user to complete the experience without having to download an app OR Lightweight and fast way for the user to complete the experience without downloading an app
Real-time validation
Real-time tracking and ML-powered validation of the scanning process
Edge computing
Landmark detection and background segmentation are held on the consumers’ devices, significantly decreasing processing time
Automated garment preparation
Our technology automatically extracts data from product pages and labels landmark points on clothing. Textures are adjusted to a fit model, ensuring your products are ready for the try-on
Natural and realistic try-on
Body proportions are preserved, products tones are adjusted to the photo, even the customer’s natural hair is reconstructed to provide a virtual try-on experience equivalent to looking into a mirror
Out-of-the-box integration
Native integration with Shopify and other platforms allows us to automatically pull the catalog information and map the fit information with individual SKUs
Intelligent Body Measurement Patented Technology #US10706262
Intelligent Body Measurement Patented Technology #US10706262
Intelligent Body Measurement Patented Technology #US10706262
Intelligent Body Measurement Patented Technology #US10706262
Intelligent Body Measurement Patented Technology #US10706262
Intelligent Body Measurement Patented Technology #US10706262
Intelligent Body Measurement Patented Technology #US10706262
Intelligent Body Measurement Patented Technology #US10706262
Intelligent Body Measurement Patented Technology #US10706262
Intelligent Body Measurement Patented Technology #US10706262
Here’s how it works
01. Body Segmentation from 2 Photos
Our technology extracts key body landmarks from a customer’s photos and creates a 3D body model from which we extract 86 measurements. We extract important features of the user’s body, such as body contour, body shape, and position, detect specific body parts (head, neck, shoulders, forearms, and ankles) and capture the details of the user’s appearance, such as haircut and skin tone.
02. 3D Model Processing
Key body landmarks from the photos are processed to create a 3D avatar. The avatar allows us to capture key body shape characteristics when producing a size recommendation and allows us to suggest a larger or smaller size, depending on the product’s fit intent. The 3D avatar generation pipeline is so robust that it will even work for shoppers with older smartphone cameras.
03. Size and Fit Recommendations
We use various data sets to produce accurate size and fit recommendations. We start by collecting brand-specific fit standard data, including core body measurements (fit form, avatar, live models) and grading rules. Then we analyze season or collection-specific tech packs to understand the fit intent of each style. Our size recommendation algorithms map this data to a customer’s body data and fit preference to produce the optimal recommended size.
04. Virtual Try-On
Our algorithms generate a geometrically correct virtual try-on that alters the shape of the body. In addition to 3D body mesh enhancements and body tracking, a new cloth simulation machine learning algorithm recreates the texture of fabrics. The semantic generation module modifies the person’s segmentation map to identify the area on the body that should be covered with the target clothes and warps the clothing mask accordingly.
01. Body Segmentation from 2 Photos
Our technology extracts key body landmarks from a customer’s photos and creates a 3D body model from which we extract 86 measurements. We extract important features of the user’s body, such as body contour, body shape, and position, detect specific body parts (head, neck, shoulders, forearms, and ankles) and capture the details of the user’s appearance, such as haircut and skin tone.
02. 3D Model Processing
Key body landmarks from the photos are processed to create a 3D avatar. The avatar allows us to capture key body shape characteristics when producing a size recommendation and allows us to suggest a larger or smaller size, depending on the product’s fit intent. The 3D avatar generation pipeline is so robust that it will even work for shoppers with older smartphone cameras.
03. Size and Fit Recommendations
We use various data sets to produce accurate size and fit recommendations. We start by collecting brand-specific fit standard data, including core body measurements (fit form, avatar, live models) and grading rules. Then we analyze season or collection-specific tech packs to understand the fit intent of each style. Our size recommendation algorithms map this data to a customer’s body data and fit preference to produce the optimal recommended size.
04. Virtual Try-On
Our algorithms generate a geometrically correct virtual try-on that alters the shape of the body. In addition to 3D body mesh enhancements and body tracking, a new cloth simulation machine learning algorithm recreates the texture of fabrics. The semantic generation module modifies the person’s segmentation map to identify the area on the body that should be covered with the target clothes and warps the clothing mask accordingly.
Our Approaches
Precise human body data capture and visualization powered by a complex combination of innovative technologies
Computer Vision & Deep Learning
We use computer vision and deep learning to analyze the photos from which we acquire and process the body measurements and specific body shape data. Our advanced computer vision algorithms detect the dressed human body on photos taken with any smartphone on any background. Neural networks determine landmarks and produce a set of probability maps.
Proprietary Statistical Modeling
We use proprietary statistical modeling to generate human models of arbitrary complexities. We have a full pipeline here that goes from registering raw scans. The dataset of the raw scans is consistently growing due to our scanning lab. We also use our statistical modeling to generate synthetic data.
Machine Learning & 3D Matching
We use machine learning & 3D matching to build a unique 3D model of each scanned customer based on detected landmarks allowing us to accurately obtain human body measurements.
Our own independent technology stack
Our team has built our own independent technology stack for creating parametric models of the human body and reconstruction from photographs.
Our technologies and expertise cover a wide range of approaches: from 3D point cloud scans registration and parametric human body modeling to mobile-first neural networks for edge processing and virtual try-on algorithms. As a result, our solutions capture accurate body measurements and produce realistic avatars and virtual try-ons.
3DLOOK created its own body scanning lab called ScanMe to study the human body aiming and provide the most comprehensive set of unique body data points to fashion brands and retailers.
Why is our body scanning lab so special?
The scanner simulates 34 different photo configurations and collects additional body parameters: BMI, fat fold depth, body muscle, body fat percentage, etc.
The scanner has 4 dynamic cameras and combines more than 5 million points into a complete, highly accurate 3D model. These points are then used to train the neural networks and improve the algorithm behind 3DLOOK’s solutions.
We control processes inside the lab through the special in-house widget with a built-in data validation system that checks each step for inaccuracies. Therefore, we obtain the most accurate data of the person with minimized error.
Integration Options
Simple integration with Shopify, Shopify Plus, Magento, and Salesforce via a widget or API. Effortlessly connect your brand with YourFit and enable a wealth of tools and features!
Security and compliance at the core
GDPR Compliant
We have incorporated GDPR standards into data practices to make sure our customers, both in the EU or US feel secure when using 3DLOOK’s solutions.
SOC 2 Compliant
Our SOC 2 designation means that 3DLOOK has designed a set of internal controls, systems, policies, and procedures that meet the industry’s best practices for protecting our customers’ data.
SSL
We use SSL encryption to maintain the highest security and data protection standards. We regularly verify our security certificates and encryption algorithms to keep your data safe.
End-to-end encryption
We encrypt every photo and relevant file metadata with unique randomly generated encryption keys using end-to-end encryption. These keys are never sent to our servers in an unencrypted format. Accessing files is only possible with a user’s unique decryption key.
What are the assets needed for fit and size recommendations?
- Fit model/mannequin
- Grading rules/specs
- Body chart
What are the product requirements for virtual try-on?
To create more realistic clothes for virtual try-on, they should be prepared in the "Invisible Mannequin/Ghost Mannequin" style. During our flow, shoppers scan their bodies in A-pose. That's why sleeves should be separated from the body, and trouser legs should be pulled apart on the photo. The minimum image size should be 2048 x 2048 pixels. Filename extension - png.
Is my customer data private?
We put data privacy and security at the forefront! Customers can store their results and shop seamlessly across devices, with data protected by 2-way encryption as it’s transferred between our solution and your customers. Photos are processed in real-time on the customer’s device to minimize the transfer of sensitive data over the web and maintain complete data privacy.
