HU 
LECTURESTATION - subsystem/service - speech recording/display/transmission/processing subsystem (including support for the hearing impaired) for presentations, training, speeches on the auditorium screen in real time, continuously and characteristically
Information management and processing is a key determinant of competitiveness, viability, economy and security, now and in the future. The vast majority of our daily information is spoken information, generated in personal conversations or in our conversations and dictations on telecommunications and IT systems, but also and especially in education, training, information and dissemination of information. digital education in development.
Recognising the value of using speech information in new ways is a very important strategic issue, especially in education, training, information dissemination, for example in in the development of digital education.
Community, university and other education, training, lectures, information, conferences are a daily practice of human cooperation and information transfer, so its further development is a widespread need and opportunity in all subject areas, using state-of-the-art digital infocommunication solutions.
The ABSONIC system will also be available for the hearing impaired who can read, and training sessions and lectures can be organised for them.
The following range of products and solutions are based on a combination of in-house know-how, development and production, with freely usable infocommunication technologies. The products and solutions are available through online automatic speech recognition in more than 100 languages.
The LECTURESTATION subsystem of the ABSONIC framework provides the following services:
- speech perception: The accuracy of speech recognition depends to a large extent on the quality of the device's microphone. For recording presentations, wired or wireless (ear, head, clothing-mounted) microphones are required, which are purpose-built noise-cancelling, optimised for speech recognition, and which pick up the speaker's voice with our ABSONIC computer speech recording, speech management and speech processing system.
- speech regeneration: noise cancellation, automatic level control and echo cancellation
- speech recording: The presentation can be recorded on a Windows desktop or notebook computer. It can also be recorded using portable digital voice recorders and mobile phones.
- speech recognition, visualisation and automatic description Lecturers' speeches can also be displayed in quasi real time using the LECTURESTATION subsystem. The result of the character recognition of the speaker's speech is continuously displayed on the projector or display in the lecture room.
- speech recognition
- Speech recognition and display with a delay of a few seconds (5-20), as speech recognition technology tests word order probability, which requires at least four words to be spoken beforehand, so speech recognition is not word-by-word but word-context pattern-by-word
- the speech recognition accuracy is about 90-95% depending on the environmental conditions: language/ dialect, audio/acoustics, speaker's speaking ability, microphone, PA system (if connected), speech regeneration, etc. If manual correction may be necessary after automatic transcription, but even then the transcription work is reduced to about 1/3 of the original.
- Translation: translation into another language of your choice (nearly 100 languages available) at the same time as the speech is recognised
- speech and translation visualisation
- the possibilities of displaying the speech-recognised text and its translation together or separately, even just the translation:
- computer displays
- in your mobile phone browser
- TV, LCD or LED video wall
- projector
- dual projector (e.g. EPSON)
- holographic display
- other
- formats for displaying speech-recognised text and its translation
- RTF in text editor
- In HTML
- OVERLAY to any surface, even with a translucent backing
- real-time display solutions for monolingual and/or bilingual speech-recognised texts in connection with the presentation
- on a separate projector or display other than the presentation
- with the presentation projector, on the same surface as the presentation, below or above the presentation, with a software video mixing solution, even if the text is displayed below the presentation in one language and above the presentation in the other
- WiFi streaming: display of the recognised text and its translation in the browser of the computers and smartphones connected to Wifi in the auditorium.
- upload to server: storage of the speaker's speech, the recognised text and its translation on a pre-configured server in the cloud (e.g. Google Drive), from where the presentation can be accessed by the public at any time, depending on the rights of the participants or freely
- webcasting: live streaming of the performance and its transcripts via the internet during the performance.
- video subtitling: video subtitling with text in one or two languages
- document management: management options for transcription and translation of real-time or stored audio recordings
- in a common RTF document, in alternating language from line to line
- in two separate RTF documents, separated by language
- self-translation: translation of text documents not only from speech recognition but also from other sources into around 100 languages or dialects
- text search in recordings: you can search for keywords in the recordings (transcripts). The more precise the transcript, the more precise the search.
- distance learning utilisation: The solution can also be used in distance learning, as the audio recording and the recognised text of the performer can be transmitted over the Internet.
EN
