We use essential cookies for the proper functioning of the website and additional ones to make interaction with the site as convenient as possible. It helps us personalize your user experience as well as obtain analytical information to improve the service.
If you agree to accept all cookies, click "Accept all"; if not, click "Only essential". To learn more, view the Cookie Policy.
Chinese scientists have tested an algorithm that recognizes gestures
Chinese scientists have managed to develop special algorithms that allow you to control a computer using gestures. Only the upper body is involved in controlling the device.
The disadvantage of existing algorithms is that they cannot accurately capture the user's gestures and have a low speed. In fact, during movements, the car simply does not keep up with the gestures of a person.
The development of the researchers is aimed at the classification of hand types, as well as the use of gesture images from libraries and their recognition. The ratio of the area of the hands and the selection of the most frequently used positions of the palms are taken into account.
The use of new technology will reduce the number of calculations and resource base. It will also increase the recognition speed without sacrificing accuracy.
When developing the algorithm, the researchers recruited 40 volunteers, who imitated 9 machine gestures 4-5 times. And the same number of people who later duplicated them. The test results showed that the algorithm is capable of responding to gestures.
The recognition accuracy is fixed at 93%. And even in the case of adaptation - scaling or hand movement.
Chinese scientists plan to continue improving the performance of the algorithm with complications - in poor lighting conditions and with an increased number of gestures.
On the OXO website in the section "Technique" you can find devices that allow you to perform both simple actions for typing text and devices for virtual reality.
The disadvantage of existing algorithms is that they cannot accurately capture the user's gestures and have a low speed. In fact, during movements, the car simply does not keep up with the gestures of a person.
The development of the researchers is aimed at the classification of hand types, as well as the use of gesture images from libraries and their recognition. The ratio of the area of the hands and the selection of the most frequently used positions of the palms are taken into account.
The use of new technology will reduce the number of calculations and resource base. It will also increase the recognition speed without sacrificing accuracy.
When developing the algorithm, the researchers recruited 40 volunteers, who imitated 9 machine gestures 4-5 times. And the same number of people who later duplicated them. The test results showed that the algorithm is capable of responding to gestures.
The recognition accuracy is fixed at 93%. And even in the case of adaptation - scaling or hand movement.
Chinese scientists plan to continue improving the performance of the algorithm with complications - in poor lighting conditions and with an increased number of gestures.
On the OXO website in the section "Technique" you can find devices that allow you to perform both simple actions for typing text and devices for virtual reality.