TECHNOLOGY
ManoMotion’s Groundbreaking AI Framework for 3D Hand Tracking & Gesture Control
AI FRAMEWORK
Human-level vision meets intelligent interaction
Performance
+99 % accuracy, <10 ms latency
Unique IP
AI-driven gesture search engine, trained on millions of hand models.
Compatiblity
Works on standard hardware
Market Leadership
First certified vision-AI for machine safety, optimized for XR
ADVANCED HAND TRACKING
Accurate hand motion analysis in 3D space
- Tracks x, y, z rotation for joints or full skeleton
- Identifies left/right hand orientation
- Determines hand side and joint orientation
- Measures realtive depth based on distnace (0-1 range)
SKELETON TRACKING
Skeleton tracking contains the confidence and joint information of the skeleton. The joints x, y and z are provided with individual positions for each of the 21 joints of the hand skeleton. SDK can be configured to return the joint information in 2D or 3D. SDK offers one and two hand tracking.
HAND SEGMENTATION
Capable of extracting the hand texture and placing it in 3D. It is done by the addition of depth estimation that enables developers to provide experiences in which the hand occludes virtual objects and vice versa.
TRY-ON FEATURES
Enables AR developers to develop ring and watch try on solutions for the evolving e-commerce fashion industry in a very quicker and easier way.
GESTURE ANALYSIS
Understand user intent through AI-powered gesture interpretation.
- Analyzes sequences of hand movements across video frames
- Classifies gesture types based on motion patterns
- Supports experience design by categorizing user intent
- Enables behavior customization and integration in Unity
MANOCLASS
ManoClass represents the raw detection of the hand in every frame. The result is independent of the previous information as it is a per-frame analysis. For each frame given, the SDK will return a ManoClass result which can be any of the following hand classes (Grab, Pinch, Point, NoHand).
CONTINOUS GESTURES
The Continuous Gestures are a convenient package of Gesture Analysis aimed to understand and categorize if the user is continuously performing a given Gesture. Continuous Gesture is calculated based on present and past information. This means that a Continuous Gesture will be fired only if the user has kept the same hand pose for a certain amount of frames (time).
TRIGGER GESTURES
The Trigger gestures are one time gestures. These types of gestures are specific sequences of ManoClasses & hand states that when performed sequentially they will be recognised as a trigger/event, similar to a mouse click.