FAQ Technology

What channels can I use to upload material?

There are three ways to upload files to Verbit.

Upload files through API – Our open API can be customized to your needs. It’s seamless for you to deploy and will enable you to transfer a large number of files easily. In the API, we define profiles which enable you to send files with different turnaround times, different guidelines and with the option to add notes.
For more information on how to use Verbit API, visit our online API documentation in this link: https://verbitaiv3.docs.apiary.io

Manual upload – You can copy and paste any link, or simply upload files from your computer, Dropbox, Google, YouTube, etc.
Click here to go to How do I upload files manually?

Upload through Connector – You can connect to the third-party platform you use to manage your content, such as Panopto, Kaltura, Brightcove, etc. You simply retrieve files from that platform by logging in and your captions are sent back once completed. Click here for more information.

TRM Support

A typical deposition or court hearing will often take a full day (~8 hours). We have traditionally broken this up into chunks causing consistency problems. Our new enhancement provides a solution for our legal customers to edit these long files easily and quickly. In addition, we now support the transcription of long files without any inconsistencies by using a smart dynamic split and entity sharing as standard within our operations platform. For our legal market the TRM format is a standard audio file format used for capturing the audio of different speakers in a deposition or multi-party litigation scenario. Read more.

How are subtitles, closed captions, and Subtitles for the Deaf and Hard of Hearing (SDH) different from each other?

Subtitles are designed for hearing users, as they only cover spoken text and do not include sound effects or other audio elements. SDH, on the other hand, are designed for those who are deaf or hard of hearing by including additional information, such as speaker tags, sound effects and other elements outside of the speech itself. Closed captions are required by law on all public broadcasts, as per FCC regulations, and are typically formatted as white text on a black background that can be positioned anywhere on the screen.

How does Verbit’s smart technology make a difference?

Verbit incorporates the latest advancements in deep learning, neural networks and natural language understanding into an adaptive learning cycle that trains our algorithm to improve accuracy over time.

What are the advantages of Verbit’s technology versus other transcription services?

Most automatic captioning and transcription solutions fall short on meeting unique industry and customer needs. Our proprietary automatic speech recognition technology, Captivate™, for example, is trained with dedicated models that are designed with customer input for term boosting, proactive research and formatting needs. Captivate is trained on domain specific needs with a dynamic domain dictionary that is continuously updated before and during its use. Our technology delivers a bespoke solution at scale with any level of customization necessary, which isn’t possible for generic automatic speech recognition technologies.

What is Verbit’s ASR and how does it work?

Verbit’s proprietary adaptive ASR technology is based on three models: linguistic, acoustic and contextual events. It is specifically trained for the domain of the customer. This strategy produces a highly accurate initial transcript (up to 90%) within minutes of the file being submitted.

How does Verbit’s solution work?

Your files are uploaded or shared via API or a platform integration. Our in-house Automated Speech Recognition (ASR) technology then automatically transcribes the file. When using Verbit’s Pro solution, the file is then edited and reviewed by professional transcribers. Adaptive algorithms enable the technology to continually become smarter, faster, and more accurate over time.

Does Verbit have an API?

Yes. For more information on our documentation, please contact support@verbit.ai. Read more

What file formats does Verbit accept?

Verbit accepts a variety of file formats, including MP4, MPG, MP3 and more. Learn more on our support site.

Is there a minimum duration for files Verbit accepts?

Yes. We accept files that are one minute and above. Read more

Does Verbit support files with difficult audio?

Yes. Verbit’s technology includes various modes that work together to convert the audio to text, including an acoustic model that reduces background noise and echo and cancels out factors that decrease audio quality. We encourage our customers to follow best practices when recording audio for transcription, but we have technology that works to improve the output as well.

Does Verbit offer translation services?

Yes. Verbit offers a variety of translation services, including machine and human translation options in dozens of language pairings. Our expert translation services work to identify nuances to help capture translations accurately.

How does artificial intelligence fit into the process?

Artificial intelligence automates the transcription and captioning process. Once complete, any corrections or modifications to the technical output are fed back to the ASR engine through adaptive algorithms. New data is also inputted to provide context and the system is trained to recognize different accents and audio quality characteristics of our customers. These features make our technology more accurate over time.

FAQ Technology

Filters