Adobe Beefs Up Its Speech to Text
Adobe today introduced redesigned Creative Cloud video tools, including much faster speech-to-text transcription in Premiere Pro.
New features in the latest release of Premiere Pro dramatically accelerate editorial workflows, including updated speech-to-text that can now be used offline, generating transcriptions up to three times faster on Intel Core i9 and Apple M1 systems and two times faster on other processors.
With downloaded and locally stored language packs, editors can now use Speech to Text on device, leveraging the GPU and ML capabilities of modern chipsets for faster performance. All 13 supported language packs can be installed within Premiere Pro or via Creative Cloud Desktop, with the English language pack automatically installed.
"Speech to Text in Premiere Pro is already the industry's best workflow for captioning videos. With this release, it is faster and more flexible. Downloadable language packs allow the user to transcribe videos without an internet connection and offer the same exceptional accuracy that users value in the original release of Speech to Text," Adobe said in a blog post on its website.
Other additions in Premiere Pro include Remix, a capability that lets users sample, edit, and add music clips to their video content, support for more formats and device types, and performance upgrades that make the entire program work faster.
Using Adobe Premiere Pro's Speech to Text, Crooked Media introduces captioning on popular podcasts.