AI to recognize illegal voice


Alibaba has developed an AI system that can recognize illegal information such as pornography and spam in voices, reports Science and Technology Daily.
The AI voice system, which can recognize multiple languages and also local Chinese dialects, has been put online for a public test, according to the report.
Technicians with Alibaba say the AI system is able to sort through the content in voice messages and other audio on the internet and determine whether something is based on either spam or pornography.
The technology also has the ability to adapt and 'learn' through constant use.
Many online platforms in China have been found spreading pornography through Autonomous Sensory Meridian Response (ASMR), according to a notice by the National Office against Pornography and Illegal Publications in June.
The ASMR is said to be helpful in getting people to relax and sleep.
- World's first in-situ diameter-adjustable tunnel boring machine rolled out
- A date with Shandong: Adventures in Confucius' hometown
- Avatar meets reality
- Drone show lights up night sky with characters from classic Chinese film
- Shanghai police crack down on illegal drones
- Liu Haixing named head of International Department