A booth for YITU Technology at the 1st Digital China Summit in Fuzhou. Photo: Li Qiaoyi/GT
Chinese artificial intelligence (AI) unicorn YITU Technology has made inroads into the area of voice recognition, a senior company executive said on Monday, which could pit the face recognition major against China's Siri-like iFlyTek.
"The current lineup of products powered by voice recognition technology out there in the market remains far from perfect," as shown by persistent high error rates found with various speech recognition software, Wu Min, vice president of business technology at YITU, said in an interview with the Global Times on the sidelines of the first Digital China Summit in Fuzhou, East China's Fujian Province.
The three-day summit, which ends on Tuesday, has drawn many domestic companies with expertise in AI, robotics and other emerging sectors that comprise the country's burgeoning digital economy.
The low accuracy rates, among other weaknesses, can be attributed to the shortcomings of basic algorithms applied in voice recognition systems and the overall poor standard of natural language understanding, Wu said.
The Shanghai-based company, like Face++ and SenseTime, has made a name for itself as a domestic face recognition heavyweight. Meanwhile, iFlyTek, based in Hefei, East China's Anhui Province, is arguably the foremost provider of voice recognition technology in the country.
The foray of YITU into an area where the likes of iFlyTek have prevailed could blur what's seen as a clear line between two popular AI technologies, industry watchers said.
"There is some overlap between face recognition and voice recognition," Wu said. He contended that his company, having proven its algorithm capability in various case scenarios using face recognition technology, can make a genuine difference in the voice-activated segment of China's booming AI landscape.
He didn't disclose any details regarding the company's voice recognition commitment.