Snapchat Adds SoundHound’s Automatic Captioning to Videos
Snapchat videos will automatically produce captions using SoundHound’s Houndify voice engine. The new feature deepens SoundHound’s connection with Snap Inc. that began in mid-2020 when Snapchat used SoundHound for its Voice Scan filter search.
SoundHound is lending its real-time speech recognition to the social media app to produce captions for videos. Snapchat users have had to create captions for their videos manually until now. When a Snapchat user is done recording their video, they can tap the quote icon in the toolbar on the right side of the screen to implement the automatic captions and edit what they say or how they look when the video plays. The new feature continues the trend among social media platforms toward automated captioning. Twitter, TikTok, and YouTube all already have a variation on the tool. Improved speech-to-text technology and reduced concern over captioning eating up computing resources have made automated captioning cheaper and more appealing, especially when considering accessibility issues for people with impaired hearing.
“We are thrilled to help Snapchat innovate with auto captions, making content accessible to more people,” SoundHound products vice president James Hom said. “Our ongoing partnership to provide a variety of voice AI solutions is giving Snapchatters hands-free convenience, more options for creating and sharing content, and the ability to stay on-the-go while using the platform.”
Snapchat has explored building its own voice tech a few years ago with lenses that would activate during a video recording when the app detected a specific word. For instance, a filter might start filling the screen with hearts when the user says ‘love.’ Snap turned to SoundHound to move beyond that limited speech recognition in favor of the more sophisticated Houndify AI in 2018. Snap rolled out the Voice Scan feature that June, embedding the Houndify speech engine in Snapchat’s Scan platform for finding lenses. Users could activate Voice Scan by saying, “Hey Snapchat” (a wake word function since removed) and describing the filter they want. The vocal description and the camera’s direction tell Snapchat what visual and musical filters to pull from its library. SoundHound and Snap extended their relationship last year by embedding the Voice Scan feature into the second generation Spectacles smart glasses designed for the Lens creator community.
The growing partnership comes Snap Inc. reported its first-ever profitable quarter. SoundHound will join Snap in going public this year after merging with special purpose acquisition company (SPAC) Archimedes Tech SPAC Partners Co. The deal values SoundHound at $2.1 billion and is expected to bring $244 million in gross proceeds. SoundHound has also been piling up significant new deals lately. Chip giant Qualcomm is adding Houndify to its Snapdragon platform, and Netflix is incorporating a Houndify-powered voice assistant into the streaming service’s Da Vinci Reference Design Kit (RDK).