LIPSense™ 3D Body Pose Standard version provides developers with two flexible tracking modes, Lite Mode and Full Mode, allowing optimization for different hardware capabilities, latency requirements, and deployment scenarios. It is engineered for embedded AI devices, industrial applications, motion analytics, and interactive systems that require reliable, real-time human motion tracking and analysis.

Premium Experience

1. Intelligent Skeleton Logic

   Our refined topology logic prevents keypoint misconnections, ensuring that arms, legs, and torso stay correctly mapped even during complex, high-speed movements.

2. Robust Hand Tracking

   Our improved hand keypoint prediction maintains a lock on users even when they step back, expanding the usable interaction zone for large-scale environments.

3. Ultra-High Frame Rate

   A 70% boost, reaching 50 FPS for real-time precision. Our accelerated processing engine drastically delivers instant visual feedback that keeps pace with fast-moving machinery and dynamic human interactions.

4. Jitter-Free Stability

   We’ve eliminated the "shaky" look common in raw tracking data. By drastically reducing keypoint jitter, your avatars and UI elements remain rock-solid, providing a premium feel and reducing eye strain for users.

Key Features (Standard Version)

1. Flexible Tracking Modes

   The Standard Version allows developers to switch between Lite Mode and Full Mode, depending on system resources and application needs.

   • Lite Mode — High-Efficiency Full-Body Tracking (18 Joints)

     Designed for resource-constrained hardware such as embedded systems, edge AI devices, and compact PCs.

     Lite Mode (18 points): Includes eyes (2), nose, ears (2), neck, shoulders (2), elbows (2), wrists (2), hips (2), knees (2), ankles (2).

     Hardware Acceleration: Supports automatic hardware acceleration through Intel® OpenVINO™, NVIDIA® TensorRT™, and NVIDIA® CUDA®.

   • Full Mode — Comprehensive Full-Body Tracking (32 Joints)

     Provides a more detailed skeletal model suitable for advanced analytics and immersive applications.

     Full Mode (32 points): Includes eyes (2), ears (2), nose, head, neck, clavicle joints (2), shoulders (2), elbows (2), wrists (2), hands (2), fingertips (2), thumbs (2), pelvis, chest, navel, hips (2), knees (2), ankles (2), feet (2).

     Acceleration Requirement: Supports automatic hardware acceleration through Intel® OpenVINO™, NVIDIA® TensorRT™, and NVIDIA® CUDA®.

2. Multi-Person Tracking

   Supports up to 43 people tracked simultaneously.

3. Captures High-Difficulty Movements

   Capable of robustly detecting movements such as swinging and staggering.

4. Developer-Friendly
   • Programming Languages: C / C++ / C# / Python / Unity
   • Operating Systems: Windows / Ubuntu / NVIDIA® JetPack

Specifications

*The illumination settings for each 3D sensor varies. Refer to the sensor’s official documents for details on illumination settings.

Software

System Requirements

Licensing Model

• SDK Free Trial:

  The SDK is offered for software development, testing and integration. Each run is limited to 20 minutes.

• Runtime (Required for Deployment):
  • Each Runtime license is bound to one unique camera serial number.
  • Perpetual License
  • One Runtime is required per deployed camera.

Purchase Process

1. Once your order is placed, you will receive an order confirmation email.
2. Once your payment via PayPal or bank wire is confirmed, a separate email containing a link to our software licensing management system will be sent to you to guide you through registration and software download.

Upgrade Policy

• Free minor version updates within one year of purchase.
• After one year, optional annual upgrade service:
  • Standard Edition: USD 49 / year
  • Professional Edition: USD 79 / year

Use Cases

• Motion Capture Prototype
• Worker Safety and Analytics
• Immersive Virtual Interactions