Derivative works — US Patent 10482517

Defensive Disclosure for US Patent 10482517

This document outlines derivative variations of the core inventive concepts disclosed in US Patent 10482517. The purpose of this defensive disclosure is to establish prior art, thereby rendering future incremental improvements or obvious modifications by third parties as non-novel or obvious under 33 U.S.C. §§ 102 and 103. The focus is on technical enabling descriptions, specific terminology, and architectural diagrams.

The core independent claims of US Patent 10482517 generally describe:

System Claim: A system comprising virtual-commerce-environment servers and client computing platforms with imaging devices, configured to provide a virtual-outfitting interface. This interface presents selectable virtual-wearable items (representing real items) and generates a real-time or near real-time composite video feed that overlays selected virtual items onto a live video feed of the user, making the user appear to wear the real items. This system includes modules for interface, motion capture, item search/selection, and composite imaging.
Method Claim: A method performed by one or more processing devices for allowing a user to simulate wearing real-wearable items, involving steps of providing a virtual-outfitting interface with an item-search/selection portion, allowing user selection of virtual-wearable items, and providing a main display portion with a composite video feed of the user and selected virtual items.

The following derivatives expand upon the system claim, as variations in system architecture and components directly imply corresponding method variations.

Derivative Variations

1. Material & Component Substitution

Derivative 1.1: Multi-Modal Sensor Integration for Enhanced Body and Garment Modeling

Enabling Description: The existing imaging device (e.g., camera 107) is substituted with a multi-modal sensor array comprising a structured-light 3D scanner (e.g., Intel RealSense D435i or L515 LiDAR depth camera) co-integrated with a high-resolution RGB camera and a thermal imaging sensor. The 3D scanner captures precise point-cloud data of the user's body, enabling accurate avatar reconstruction and real-time body pose estimation, replacing or augmenting the motion-capture module's (116) reliance on 2D image analysis. The thermal sensor detects heat signatures, aiding in distinguishing the user's body from background and improving segmentation robustness, particularly in challenging lighting conditions. The composite-imaging module (120) is enhanced to utilize the 3D depth map for occlusion culling and physically based rendering (PBR) of virtual-wearable items, ensuring they conform to the user's 3D geometry and appear behind or in front of body parts as appropriate. Furthermore, virtual items can be rendered with simulated material properties (e.g., reflectivity, translucency) derived from PBR models that interact with a synthetic lighting environment estimated from the live RGB feed.

graph TD
    A[Client Computing Platform 106] --> B{Multi-Modal Sensor Array};
    B -- Live RGB Video --> C(Motion-Capture Module 116);
    B -- 3D Point Cloud Data --> C;
    B -- Thermal Data --> C;
    C -- Body Pose & Geometry --> D(Composite-Imaging Module 120);
    E[Item-Search/Selection Module 118] -- Selected Virtual-Wearable Item (PBR Model) --> D;
    D -- Real-time 3D Composite Feed --> F[Virtual-Outfitting Interface 114];
    F --> G[User Display];

Derivative 1.2: Haptic Feedback Integration via Wearable Actuators

Enabling Description: To enhance the user's immersive experience, the client computing platform (106) is augmented with wearable haptic feedback devices (e.g., wristbands, vests, or full-body suits equipped with linear resonant actuators or eccentric rotating mass motors). When a virtual-wearable item (selected via item-search/selection module 118) is overlaid on the user, the system's processor (112), specifically through an extended interface-control module (130) or a new haptic feedback module, sends actuation signals to the haptic devices. These signals simulate sensations associated with the virtual garment, such as the texture of fabric (e.g., rough denim, smooth silk), the weight of an accessory, or the subtle pressure of a fitting item. The haptic feedback is dynamically mapped to zones on the user's body corresponding to the virtual item's placement, with intensity and frequency modulated based on the virtual item's characteristics (e.g., "type of fabric, texture of fabric" as mentioned for motion-capture module 116).

graph TD
    A[Client Computing Platform 106] --> B[Virtual-Outfitting Interface 114];
    B -- Selected Virtual-Wearable Item --> C(Haptic Feedback Module);
    C -- Actuation Signals --> D[Wearable Haptic Devices];
    D -- Tactile Feedback --> E[User];
    F[Composite-Imaging Module 120] -- Visual Feedback --> B;

Derivative 1.3: GPU-Accelerated Real-time Cloth Physics Simulation

Enabling Description: The virtual-commerce-environment servers (102) or client computing platforms (106) incorporate a dedicated GPU (Graphics Processing Unit) accelerator, such as NVIDIA GeForce RTX series or AMD Radeon RX series, specifically utilized by the composite-imaging module (120). This GPU is employed for real-time cloth physics simulation, departing from simple visual overlay. The selected virtual-wearable item is represented by a high-polygon mesh with material properties (e.g., elasticity, stiffness, damping coefficient) derived from the "characteristics associated with a virtual-wearable item." The motion-capture module (116) provides not just body position and orientation, but also skeletal animation data. The GPU-accelerated cloth simulation engine (e.g., using CUDA, OpenCL, or Vulkan compute shaders) processes these inputs to simulate the realistic drape, fold, and movement of the virtual garment over the user's body in response to the user's motions, gravity, and simulated wind effects. This significantly enhances the "realness of the appearance of the virtual-wearable item" beyond static overlays.

graph LR
    A[Motion-Capture Module 116] -- Skeletal Animation & Pose --> B(GPU-Accelerated Cloth Physics Engine);
    C[Item-Search/Selection Module 118] -- Virtual Garment Mesh & Material Properties --> B;
    B -- Simulated Garment Geometry --> D(Composite-Imaging Module 120);
    D -- Composite Video Feed --> E[Main Display Portion 204];

2. Operational Parameter Expansion

Derivative 2.1: Ultra-High-Resolution (16K) Volumetric Video Streaming for Virtual Try-On

Enabling Description: The system is adapted to operate with ultra-high-resolution volumetric video. Instead of a standard 2D or 3D camera (107) capturing video, a volumetric capture studio, comprising multiple synchronized 4K depth and RGB cameras (e.g., Azure Kinect sensors or specialized volumetric capture arrays), captures the user. This generates a 16K-equivalent resolution volumetric video stream (point cloud + texture) of the user at 90 frames per second. The client computing platform (106) and virtual-commerce-environment servers (102) are equipped with high-performance network interfaces (e.g., 100 Gigabit Ethernet) and specialized hardware decoders (e.g., FPGA-accelerated H.266/VVC or V-PCC decoders). The motion-capture module (116) processes this dense volumetric data for highly accurate body segmentation and articulation, even detecting subtle muscle flexes. The composite-imaging module (120) performs real-time volumetric rendering, seamlessly integrating volumetric virtual-wearable items into the user's volumetric representation, allowing for realistic depth interaction and parallax effects viewable from multiple angles, exceeding typical flat overlays.

graph TD
    A[Volumetric Capture Studio (Multi-4K Cams)] --> B(Volumetric Video Encoder/Streamer);
    B -- 16K Volumetric Stream (90fps) --> C[High-Performance Network];
    C --> D[Client/Server (Hardware Decoder)];
    D --> E(Motion-Capture Module 116 - Volumetric);
    D --> F(Composite-Imaging Module 120 - Volumetric);
    E -- User Volumetric Pose --> F;
    G[Item-Search/Selection Module 118] -- Volumetric Virtual Item --> F;
    F -- Real-time Volumetric Composite --> H[Virtual-Outfitting Interface (Multi-view)];

Derivative 2.2: Hyperspectral Imaging for Material Property Extraction and Simulation

Enabling Description: The imaging device (camera 107) is replaced with a hyperspectral imager (e.g., a push-broom scanner with a spectral range of 400-1000 nm, capturing 200+ spectral bands). This imager captures detailed spectral reflectance signatures of the user's existing clothing or skin. The motion-capture module (116) is extended to include a spectral analysis sub-module that, for example, identifies the type of fabric, texture, color, and even subsurface scattering properties of the user's current attire. This information is then used by the composite-imaging module (120) to dynamically adjust the rendering of the virtual-wearable item. For instance, if the user is wearing a dark, matte shirt, and the virtual item is a reflective metallic accessory, the composite rendering will accurately reflect how the virtual item would appear against the user's actual material properties, improving color consistency and lighting interaction, enhancing the "realness."

graph TD
    A[Hyperspectral Imager] --> B(Spectral Data Acquisition);
    B --> C(Spectral Analysis Sub-module);
    C -- Material Properties (User) --> D(Composite-Imaging Module 120);
    E[Item-Search/Selection Module 118] -- Virtual Item (Material Properties) --> D;
    D -- Spectrally Enhanced Composite --> F[Main Display Portion 204];

3. Cross-Domain Application

Derivative 3.1: Surgical Instrument Simulation (Medical Field)

Enabling Description: The system is adapted for simulating the "wearing" or "holding" of surgical instruments and prosthetics in a medical training or planning context. The "virtual-wearable items" are replaced with virtual surgical tools (e.g., scalpels, forceps, endoscopes) or prosthetic devices. The imaging device (107) captures a live video feed of a trainee surgeon or a patient's anatomical area (e.g., a mannequin or a pre-operative scan projected onto a physical model). The motion-capture module (116) tracks the trainee's hands and fingers, or specific anatomical landmarks on the physical model. The composite-imaging module (120) overlays the virtual surgical instruments onto the trainee's hands or the virtual prosthetics onto the anatomical model, in real-time. This allows for virtual practice of surgical procedures or visualization of prosthetic fit without physical contact, providing "real-time or near-real time" feedback on positioning, orientation, and interaction. The "purchase module" (124) could be adapted to "select" instruments for a virtual tray.

graph TD
    A[Imaging Device (Surgical Field)] --> B(Live Video Feed of Trainee/Anatomy);
    B --> C(Motion-Capture Module 116);
    C -- Hand/Anatomy Tracking --> D(Composite-Imaging Module 120);
    E[Item-Search/Selection Module 118 (Surgical Tools)] -- Selected Virtual Surgical Tool/Prosthetic --> D;
    D -- Real-time Composite (Tool Overlay) --> F[Virtual-Outfitting Interface (Surgical Training)];
    F --> G[Trainee Display];

Derivative 3.2: Industrial Safety Equipment Training (Manufacturing/Industrial)

Enabling Description: The system is re-purposed for training workers on proper donning and fitment of industrial safety equipment. "Virtual-wearable items" become virtual hard hats, safety glasses, respirators, specialized gloves, or full hazmat suits. The imaging device (107) captures a live feed of an industrial worker. The motion-capture module (116) tracks the worker's head, face, hands, and body posture to determine the correct placement of safety gear. The composite-imaging module (120) overlays the virtual safety equipment, providing a visual simulation of correct or incorrect fit. An "interface-control module" (130) enhancement can highlight (e.g., red/green indicators) areas of improper fit based on predetermined safety standards, enabling immediate visual feedback for training. The "snapshot module" (126) allows supervisors to capture and review worker's training progress.

graph TD
    A[Imaging Device (Industrial Site)] --> B(Live Video Feed of Worker);
    B --> C(Motion-Capture Module 116);
    C -- Worker Body Tracking --> D(Composite-Imaging Module 120);
    E[Item-Search/Selection Module 118 (Safety Gear)] -- Selected Virtual Safety Gear --> D;
    D -- Real-time Composite (Safety Gear Overlay) --> F[Virtual-Outfitting Interface (Safety Training)];
    F --> G[Worker Display & Feedback];

Derivative 3.3: Virtual Exhibit Design and Installation (Museum/Exhibition)

Enabling Description: The system is utilized for virtual planning and visualization of museum exhibits or art installations. The "virtual-wearable items" are re-contextualized as virtual exhibits, artifacts, sculptures, or display cases. The imaging device (107) captures a live video feed of an empty exhibition space or a scale model of it. The motion-capture module (116) tracks an exhibition designer's hand gestures or an augmented-reality marker placed within the space to manipulate the position and orientation of virtual exhibits. The composite-imaging module (120) overlays these virtual exhibits into the live video feed of the physical space, allowing designers to virtually "try on" different layouts and visualize the final appearance of an exhibition in real-time, facilitating collaborative design. The "conferencing module" (122) could allow multiple designers to collaborate virtually in the same digital space.

graph TD
    A[Imaging Device (Exhibition Space)] --> B(Live Video Feed of Room);
    B --> C(Motion-Capture Module 116);
    C -- Designer Gestures/Marker Tracking --> D(Composite-Imaging Module 120);
    E[Item-Search/Selection Module 118 (Exhibit Catalog)] -- Selected Virtual Exhibit/Artifact --> D;
    D -- Real-time Composite (Exhibit Overlay) --> F[Virtual-Outfitting Interface (Exhibit Design)];
    F --> G[Designer Display];

4. Integration with Emerging Tech

Derivative 4.1: AI-Driven Fit Optimization and Recommendation Engine

Enabling Description: The system integrates an AI-driven fit optimization and recommendation engine. The motion-capture module (116) is augmented to not only track user motion but also extract detailed anthropometric measurements and body shape parameters (e.g., shoulder width, waist circumference, limb length, torso depth) from the live video feed. This data, along with user preferences and historical purchase data (from electronic storage 110), is fed into a machine learning model (e.g., a deep neural network trained on a large dataset of body scans and garment fit data). This AI model, residing on the virtual-commerce-environment servers (102), analyzes the virtual-wearable item's 3D model and the user's estimated body shape to predict optimal sizing and fit, and suggest alternative styles or sizes. The "current item details portion" (206) is updated to display AI-generated fit scores, recommended sizes, and personalized styling advice, dynamically influencing the "item-search/selection module" (118) to present optimized suggestions.

graph TD
    A[Motion-Capture Module 116] -- Anthropometric Data --> B(AI Fit Optimization Engine);
    C[Electronic Storage 110] -- User Preferences & Historical Data --> B;
    D[Item-Search/Selection Module 118] -- Virtual Item 3D Model --> B;
    B -- Recommended Size/Style/Fit Score --> E[Current Item Details Portion 206];
    E --> F[Virtual-Outfitting Interface 114];
    B -- Refined Item Suggestions --> D;

Derivative 4.2: IoT Sensor-Enhanced Garment Data for Hyper-Realistic Simulation

Enabling Description: Real-wearable items available for virtual try-on are augmented with embedded Internet of Things (IoT) sensors (e.g., strain gauges, accelerometers, temperature sensors, fabric-property micro-sensors). These sensors continuously transmit real-time data about the garment's actual drape, deformation under gravity, interaction with air currents, and textile properties when worn by a physical model in a controlled environment. This live IoT data is wirelessly transmitted (e.g., via Wi-Fi, Bluetooth Low Energy, or 5G) to the virtual-commerce-environment servers (102) and ingested by the composite-imaging module (120). Instead of relying solely on pre-defined material properties, the composite-imaging module uses this real-time IoT data to drive a physics-based rendering engine, dynamically adjusting the virtual-wearable item's appearance (e.g., wrinkling, stretching, flow) to match the real garment's observed behavior, achieving "hyper-realness" that reacts precisely to environmental forces and gravity.

graph TD
    A[Real-Wearable Item (with IoT Sensors)] --> B(IoT Gateway/Network);
    B -- Real-time Garment Data --> C[Virtual-Commerce-Environment Servers 102];
    C --> D(Composite-Imaging Module 120);
    E[Motion-Capture Module 116] -- User Pose --> D;
    F[Item-Search/Selection Module 118] -- Virtual Item ID --> D;
    D -- Hyper-Realistic Composite --> G[Main Display Portion 204];

Derivative 4.3: Blockchain for Digital Asset Rights and Provenance of Virtual Items

Enabling Description: The system integrates blockchain technology to manage the provenance, ownership, and licensing of virtual-wearable items and snapshots. Each virtual-wearable item's 3D model, texture files, and associated metadata (e.g., designer, brand, usage rights) are registered as non-fungible tokens (NFTs) on a decentralized ledger (e.g., Ethereum, Solana, or a private blockchain). When a user selects a virtual item (via item-search/selection module 118) or captures a snapshot (via screen-capture module 126), the system verifies usage rights via a blockchain client on the virtual-commerce-environment servers (102). The "social-networking module" (128) can share these snapshots with embedded cryptographic proofs of authenticity, ensuring that the source of the virtual garment and the image's integrity are verifiable. The "purchase module" (124) can facilitate the purchase of digital ownership or limited-use licenses for virtual items or digital copies of real items, recorded on the blockchain.

graph TD
    A[Virtual-Commerce-Environment Servers 102] --> B(Blockchain Client/Node);
    B -- NFT Registration/Verification --> C[Decentralized Ledger (Blockchain)];
    D[Item-Search/Selection Module 118] -- Virtual Item Selection --> A;
    E[Screen-Capture Module 126] -- Snapshot Capture --> A;
    F[Social-Networking Module 128] -- Share Snapshot --> A;
    A -- Usage Rights/Provenance Check --> C;
    A -- Record Digital Purchase/License --> C;
    C -- Verification Data --> G[User (Authenticity/Ownership)];

5. The "Inverse" or Failure Mode

Derivative 5.1: "Privacy-Aware" Low-Resolution/Anonymized Mode

Enabling Description: The system includes a "privacy-aware" operational mode, activated via the "virtual-outfitting-interface-control tool" (130). In this mode, the imaging device (107) captures a significantly lower resolution video feed (e.g., 240p at 10 fps) and applies real-time anonymization filters via the motion-capture module (116). These filters could include facial blurring, stylization (e.g., turning the user into a generic silhouette or avatar), or body keypoint extraction without rendering the full video frame. The composite-imaging module (120) then overlays the virtual-wearable item onto this anonymized representation. This mode is designed for scenarios where users are hesitant to share high-fidelity live video, allowing for limited-functionality try-on with reduced personal data exposure, serving as a "safe failure" for privacy concerns by degrading visual fidelity but maintaining core functionality.

graph TD
    A[Imaging Device 107] -- Live Video Feed --> B(Motion-Capture Module 116);
    B -- Anonymization Filters (Blur/Silhouette/Keypoints) --> C(Composite-Imaging Module 120);
    D[Item-Search/Selection Module 118] -- Selected Virtual Item --> C;
    C -- Low-Res/Anonymized Composite --> E[Main Display Portion 204 (Privacy Mode)];
    F[Virtual-Outfitting-Interface-Control Tool 130] -- Activate Privacy Mode --> B;

Derivative 5.2: "Limited Bandwidth" Degraded Performance Mode

Enabling Description: When network conditions degrade (e.g., detected by a network monitoring sub-module within the interface-control module 130), the system automatically enters a "limited bandwidth" mode. In this mode, the client computing platform (106) reduces the resolution and frame rate of the outgoing video feed from the imaging device (107) (e.g., from 1080p@30fps to 480p@15fps). Furthermore, the virtual-commerce-environment servers (102) simplify the rendering complexity of virtual-wearable items. This simplification can include reducing polygon count of 3D models (e.g., Level of Detail (LOD) switching), using lower-resolution textures, disabling advanced physics simulations, or switching to pre-rendered static overlays rather than real-time dynamic ones. This ensures basic virtual try-on functionality remains available even under adverse network conditions, operating in a "limited-functionality" mode to prevent complete service interruption.

graph TD
    A[Network Monitor (Interface-Control Module 130)] -- Bandwidth Degradation --> B(System Control Module);
    B -- Adjust Video Settings --> C[Imaging Device 107];
    B -- Simplify Rendering --> D(Composite-Imaging Module 120);
    C -- Reduced Quality Video --> D;
    E[Item-Search/Selection Module 118] -- Simplified Virtual Item --> D;
    D -- Degraded Performance Composite --> F[Main Display Portion 204 (Low Bandwidth Mode)];

Combination Prior Art Scenarios

US10482517 + OpenCV (Open-Source Computer Vision Library):
The motion-capture module (116) as described in US10482517, which tracks user motion, recognizes body parts, and determines position/orientation for virtual item placement, could be implemented using standard image processing and computer vision algorithms available in the open-source OpenCV library. Specifically, functionalities such as background subtraction, facial recognition (e.g., Haar cascades), body pose estimation (e.g., OpenPose or MediaPipe integrations), and augmented reality marker detection (e.g., ArUco markers) are well-documented and widely accessible within OpenCV. Combining the high-level system and method claims of US10482517 with known, publicly available, and open-source computer vision techniques for gesture detection and object tracking would render many specific implementation details of motion capture and virtual item positioning obvious to a person having ordinary skill in the art.
US10482517 + Three.js (Open-Source JavaScript 3D Library):
The composite-imaging module (120) and the virtual-outfitting interface (114) of US10482517, particularly for web-based applications, could leverage the open-source Three.js library. Three.js provides a high-level API for creating and displaying animated 3D graphics in a web browser using WebGL. A person skilled in the art could combine US10482517's concept of overlaying virtual-wearable items onto a live video feed by using Three.js to render the 3D virtual garments and then compositing this 3D scene with a live video stream (e.g., from getUserMedia() API) in real-time within an HTML5 canvas element. Three.js's capabilities for loading 3D models (e.g., GLTF, OBJ), applying textures, and handling lighting would directly enable the "rendering the virtual-wearable item in the main display portion in order to enhance the realness of the appearance."
US10482517 + WebRTC (Open-Source Standard for Real-time Communication):
The conferencing module (122) described in US10482517, which allows a user to "view and interact with friends that are also virtually trying on real-wearable items" and to "interact, such as by voice or text, with other users," can be readily combined with the open-source WebRTC standard. WebRTC allows web browsers and mobile applications to communicate directly peer-to-peer, facilitating real-time video, audio, and data exchange without requiring plugins. Implementing the conferencing portion (208) by using WebRTC for live video/audio streams and data channels would be an obvious choice for a skilled practitioner. This combination would enable the "enhanced sense of shopping with other users in disparate locations" envisioned by the patent, as WebRTC handles the complexities of media capture, codec negotiation, and peer-to-peer connectivity, abstracting away the underlying networking infrastructure.