Roku Applies Speech Recognition AI to Make Too-Quiet Dialogue Audible
Roku has deployed its speech recognition AI to solve the issue of people speaking too quietly in movies and television to be heard. The new Roku OS 11 incorporates several voice AI improvements and additions, including the new automatic speech clarity feature to detect when people are talking and raise the volume of the dialogue without making the rest of the audio louder.
The automatic speech clarity setting operates much like the noise-cancellation tech in some phones and digital communications platforms The AI’s automated speech recognition model differentiates between humans voices and other sounds, isolating the portion of the audio to boost. Roku’s AI raises the volume of the voices without reducing or eliminating the non-voice sounds as it would on the phone. People who feel compelled to use subtitles because the people onscreen seem to be whispering over too-loud music will likely be eager to deploy Roku’s AI for that reason.
“At Roku, we are laser-focused on tailoring the streaming and visual experience to fit our customers’ personal preferences,” oku senior vice president of product and experience Gidon Katz said. “We recognize not everyone interacts with their TV in the same way, so we are proud to offer a platform full of choice, while simultaneously providing our users an easy-to-use experience. With Roku OS 11, we’re offering a platform with new personalized updates across search, audio, content discovery along with a new feature that allows our customers to display and even share photo albums through Roku devices.”
Speech clarity isn’t the only additional AI feature in Roku OS 11. The new Sound Modes, adjust audio controls for different kinds of content, including dialogues, music, and movies. The Roku voice assistant added a keyboard control last year to search for content and enter logins and passwords, but only in English. The updated OS adds Spanish, German, and Portuguese voice keyboards to the list. Roku’s AI is a key element as it competes with a growing number of smart TV rivals with their own strategies and AI features. For instance, TCL recently picked Disruptel to add a voice assistant to its TVs. Amazon’s Fire TV platform started a renewed voice emphasis using new Amazon-produced Fire TVs. Meanwhile, Google has plans to make TVs the center of its smart home ecosystem. SoundHound recently struck a deal with Vizio to incorporate SoundHound’s Houndify voice assistant platform into most Vizio smart TVs. The company has also teamed with Netflix to add a Houndify-powered voice assistant into the streaming service’s Da Vinci Reference Design Kit (RDK).