 Interaction
with eye contact. |
"Wakamaru" knows many different words.
It will recognize its own name as assigned by the owner, respond when spoken to, and offer up subjects of conversation itself.
The face recognition feature allows it to differentiate the names of up to 100 people.
During a conversation, it will direct its body towards the speaker, make eye contact, and also use gestures. |
 Interactive voice technology (voice recognition and speech technology) |
"Wakamaru's" voice recognition engine is called "Word Spotting." Although word spotting does not allow it to understand long sentences by itself like a human, "wakamaru" can recognize pre-determined single words.
"Wakamaru's" conversation skills follow the rules of a constant scenario by recognizing single words and responding accordingly.
Also, "wakamaru" is equipped with the latest speech engine. Furthermore, it is able to fluently read aloud any information in real time, such as the news or weather forecast, which could not be programmed ahead of time. |
| Eye
contact technology (physical or facial movement detection technology) |
| "Wakamaru" is equipped with two cameras-one on its face and the other on top of its head-so that it can always be aware of "wakamaru's" surroundings. In particular, the camera on the top of its head seeks out any movement of people and faces around "wakamaru" while rotating 360 degrees, in every direction. |
360 degree omni-direction camera. |
 |
 |
 |
 |
| Turns towards movement. |
|
Seeks out faces to make eye contact. |
|
 |
|
 |
|
| Face recognition technology (face observation and differentiation technology) |
| After eye contact is established, identity is verified. By searching and verifying faces recognized using facial characteristics, "wakamaru" can accomplish tasks such as delivering mail to the correct recipient, etc. |
|
|
 Face Recognition Technology (Technical cooperation : OMRON Corporation) |
 |
| Facial Position Detection |
Facial Feature Extraction |
References Face Database |
| Features ： |
Robust to condition changes of background and lighting
Robust to changes in expressions and tilting of head |
| |
|
|
|
|
|
