AppTek Launches OmniCaption 300
AppTek, a provider of artificial intelligence, automatic speech recognition (ASR), neural machine translation, text-to-speech, and natural language processing and understanding technologies, has launched the OmniCaption 300 captioning appliance, a stand-alone server that resides on broadcasters' premises for real-time, fully automated, same-language live closed captions for broadcast content.
OmniCaption 300 accommodates more than 30 languagesand incorporates the latest advancements in neural-based ASR technology embedded with features like noise adaptation, vocabulary normalization, automatic punctuation, true-casing, and speaker separation. It also offers advanced on-device language model adaption, including options for profanity filtering, word replacements, number formatting, and a custom dictionary tool to manage non-standard, out-of-vocabulary items, such as local place names, unique terms such as Brexit or Quidditch, or lesser-known proper names.
"The quality and speed of the OmniCaption300 model shows how far the technology has come in the last couple of years," said Juan Mario Agudelo, vice president of sales for media and entertainment at AppTek, in a statement. "Any broadcast media channel, university, or local governance will be surprised at its performance, quality, and dependability as either a primary or backup solution for always-on captions."
In addition to captioning live content, the OmniCaption 300 can also be used for the transcription of media and archived content to generate subtitles or metadata for the discoverability of archived assets inside media asset management systems. The device is pre-installed with AppTek's Workbench Lite,a post-editing transcription application integrated with ASR technology and storage for offline content.