A new version of AUDIMUS MEDIA was released at NAB Show 2017 and the feedback\u00a0was really positive!\u00a0This solution is now working 24\/7 in several new clients\u00a0in Brazil, among which we highlight:\u00a0TV Serra Dourada (Goi\u00e2nia),\u00a0SBT MS (Campo Grande),\u00a0TV Cultura (S\u00e3o Paulo), Rede Vanguarda (S\u00e3o Jos\u00e9 dos Campos) and\u00a0TV Rio Sul (Resende). Also, in USA, the software is being tested by more than 50 affiliates of the main TV Networks.<\/p>\n
But what does this solution promise? It is really clear-cut: it generates subtitles for live TV broadcasts or even streamed contents, without any human intervention, while\u00a0complying\u00a0with legal requirements imposed for TV broadcasters to produce subtitles for their contents and\u00a0avoiding prohibitively high costs of manual subtitling. It delivers high accurate subtitles in real-time for any type of programming. The system’s fifth generation comprises several major improvements, such as the integration of deep-learning techniques to optimize the performance of acoustic models derived from the ever-growing speech repositories, an updated version of the American English language model with its extended vocabulary and a reduction\u00a0on latency time for flawless subtitles during broadcast.<\/p>\n
VoiceInteraction\u00a0believes that the best way for clients to fully understand the potential of a product is by testing it inhouse. That is why we have been\u00a0encouraging TV Stations to test AUDIMUS.MEDIA, without a\u00a0purchase compromise.<\/p>\n