Koemei Unveils Cloud-Based Speech Recognition and Transcription Solutions

Video and audio transcription provider Koemei has launched Koemei Web Service, a cloud-based platform and programming interface for the next generation of video transcription and captioning.

According to the company, transcription and captioning of video and audio content can now be performed automatically and more accurately, replacing costly and cumbersome manual transcription methods.

The Koemei Web Service platform enables transcription of video and audio content for captioning, indexing, search and discovery, and search engine optimization. At the core of Koemei Web Service is its speech decoding engine, which converts the audio of every speaker into text, using the cloud-based multi-speaker speech recognition platform and API for transcription of video and audio content.

Koemei Web Services also offers the following capabilities:

  • a cloud-based self-service platform and API so there's no need for technical skills in speech; users can add capabilities with simple HTTP requests using a RESTful API.
  • a wiki-type edit platform for editing by a closed or open community for accessibility compliance.
  • the ability to accept input from various sources (language model, pronunciation dictionary, acoustic models) into a single decoder for greater accuracy.

"Koemei Web Services meets the need in business, media, governments, education, and developers who produce an ever-increasing amount of video but are limited to current manual, and costly, transcription methods," said Temitope Ola, co-founder and CEO, Koemei, in a statement. "We now offer a cloud-based, fully-automated machine transcription solution for enterprises to caption, index, and monetize their content.

"Current solutions are manual and costly ranging from $2.00 to $5.00 a minute and are not suitable or sustainable for large scale transcription needs of the media, government, or education. In the U.S. alone, video and audio producers lose close to $5 billion annually through the inability of outdated manual methods to transcribe large amounts of content," Ola said.

SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues