Change the face to portray any character with diverse gender, age, and human likeness.
The groundbreaking combination of facial animation and replaceable masks make it easy and affordable to create expressive characters for any purpose.
Furhat is shipped with an adult mask and 22 animated faces that can be customized into further characters . Child, anime and dog masks are available for purchase, and additional customized faces and masks can be created on-demand.
Ingenious back-projection system to display a digital face on a physical mask
One of Furhat’s very unique features is the face. The projector inside the Furhat head projects a 3D animation onto a translucent mask to create a face that feels alive. This combination of hardware and software give the robot the flexibility of a digital animated face with the impact of having a physical presence. Using special materials, the masks are designed for maximum resolution and contrast to absorb light from an optical projection position behind the mask.
Wide selection of synthesized voices in different languages – or use real voices
Furhat offers access to more than 200 voices in over 35 languages and varying regional dialects. Choose the voice that best suits your character, whether it’s young or old, male or female. Or use real voice recordings and the robot will lip sync to it automatically.
Automatic face movements gives illusion of life
The robot has built in automatic behaviours such as blinking, micro movements of the face and eyes that provides a feeling that the face is alive. Raising the eyebrows when emphasizing a word or glancing over at another person facilitates realistic and natural interactions.
Natural facial expressions, head movements and eye gaze
The combination of natural head movements with three degrees of freedom, fluent facial gestures and eye gaze creates a highly expressive robot with a very natural behavior.
Advanced Face Engine enables highly expressive animations
The face of Furhat is animated in Unity, a state-of-the-art 3D animation engine. Face skins (textures) are created using industry standard graphical design tools together with a system of character parameters for changing the geometry of the face. The standard 3D model includes close to one hundred parameters that enables a variety of faces and highly expressive face gestures with detailed control.
Advanced computer vision for face tracking and estimation of engagement
Furhat uses computer vision to track multiple users in real-time, do facial expression analysis, estimate head pose and user distance. This includes:
- Highly accurate face detection system based on state-of-the-art deep learning (single-shot-detector) model
- Accurate multi-user tracking, depth estimation, and spatial modelling
- Supports face recognition using DNN-based face model, allows keeping track of users over time
Combined visual and audio input enables multi-user interactions
Furhat can interact with up to 10 people simultaneously and has the ability to distinguish between them and remember them during the course of an interaction.
Natural interactions by maintaining eye contact and returning smiles
Set automatic behavior for returning smiles, or get access to the facial expression to set thresholds and use the analysis to model engagement, attention or the internal state of the user. The analysis includes Happiness, Surprise, Anger, Sadness – with Happiness being the most robust facial expression to detect.
Fluid conversations with a powerful NLU engine
The Furhat Platform is designed for natural conversations with rapid turn-taking. You have detailed controls for managing initiative, turn-taking, interruptions, error handling and priming the speech recogniser for expected utterances. The integrated NLU engine supports a number of languages and lets you easily define intents and entities to extract meaning from user utterances.
Powerful speakers and array mic for noisy environments
The robot comes with an external USB microphone array with noise cancelation and direction of arrival. You can also plug in any microphone to the USB-jack in the back of the robot. Powerful stereo speakers allows the robot to be heard in noisy environments.
State-of-the-art speech recognition and TTS in multiple languages
Furhat has support for more than 120+ languages using Google Cloud Speech-To-Text and Microsoft Azure Speech-To-Text. For SpeakingTTS, Furhat is using Amazon Polly Cloud TTS as well as Acapela on-board TTS, together covering more than 200 voices and 35 languages.