What is Embody Digital?
Embody Digital offers a specialized Behavior API designed to automatically animate 3D characters, avatars, or robots based purely on text or voice input. This technology aims to imbue digital entities and robots with human-like communication capabilities, enhancing interaction realism. The system generates comprehensive animations including lip syncing, emotional expressions, head movements, saccadic eye movements, and synchronized body gestures that reflect natural human conversational patterns.
The platform facilitates the embodiment of conversational AI systems, such as those built with Google DialogFlow, IBM Watson, Amazon Lex, RASA, or any other text-based input mechanism, allowing for near real-time animation. Users can refine the animation performance by incorporating inline tags within the input text to specify emotional tone (e.g., happy, sad, sarcasm) or stress levels on particular words. The API response provides detailed behavior data synchronized with the spoken words and includes Behavior Markup Language (BML) blocks, suitable for direct use with BML realizers like SmartBody to drive the animation.
Features
- Behavior API: Automatically animates 3D characters, avatars, or robots from text or voice.
- Human-like Animation: Includes lip syncing, emotional expressions, head movements, eye movements, and body gestures.
- Emotion Tagging: Allows specifying emotional tone (happy, sad, angry, etc.) for nuanced performance.
- Stress Tagging: Enables emphasizing specific words and associated behaviors.
- Conversational AI Integration: Works with systems like chatbots, Google DialogFlow, IBM Watson, Amazon Lex, RASA.
- BML Output: Generates Behavior Markup Language for direct animation realization (e.g., via SmartBody).
- Unreal Engine Integration: Provides code samples for use with Unreal Engine and Metahumans.
Use Cases
- Animating virtual assistants and chatbots for more engaging interactions.
- Bringing digital characters in games or simulations to life.
- Creating expressive avatars for virtual reality or metaverse applications.
- Developing animated characters for educational or training software.
- Enhancing robotic interactions with human-like non-verbal cues.
FAQs
-
How does the Embody Digital Behavior API work?
It processes text or voice input, including optional emotion and stress tags, to generate timed behavior data and BML for animating 3D characters, avatars, or robots with human-like expressions and movements. -
What kind of animations does the API generate?
It generates lip syncing, emotional expressions, head movements, saccadic eye movements, body gestures, and synchronized body movements. -
Can I control the emotional expression of the animation?
Yes, you can use inline tags (e.g., [happy], [sad]) or specify emotional phrases via the API call to control the character's emotional tone. -
Does Embody Digital provide integration examples?
Yes, code samples for accessing the API using Unreal Engine 4.26 and Metahumans are available, along with dependencies like a modified SmartBody library and a TTS connector. -
How can I get access to the Behavior API?
Interested users should contact info@embodydigital.com for terms of use and licensing details for both the API service and the embeddable library.
