The popularity and importance of video call and conferencing technologies have grown hugely since the COVID-19 pandemic.
With more people than ever working from home, demand only looks set to increase.
According to figures from Upwork, 80% of employees use video calling for one-to-one meetings and 75% of CEOs believe that video conference calls will entirely replace audio-only conference calls.
Existing technology still faces challenges, however, and issues such as poor transmission quality, dropouts, and failure to connect can be very frustrating.
Now, a team has tested a method for transmitting video conference calls using extremely low bandwidth connectivity.
They did so by transmitting data on a dive to one of the most famous shipwrecks of all time – RMS Titanic.
Titanic sank on its maiden voyage in the North Atlantic after striking an iceberg in 1912.
The remains of the ship lie at a depth of nearly 4,000 metres below the surface of the sea.
Professor Alex Waibel, who conducts research on speech translation at Karlsruhe Institute of Technology (KIT) in Germany and Carnegie Mellon University (CMU) in the US, said that the act of transmitting data through almost 4km of salty water without loss was extremely challenging.
Radio communication does not work in saltwater, so transmissions can usually only be sent from a submersible to its parent vessel at surface level.
Video is translated to text and sent via sonar pulses
Researchers from KIT and CMU use synthetic techniques to convert video data into text within the submersible.
This data can then be sent to the surface via sonar sound pulses, where it can be reconstructed.
This involves mapping a synthetic voice to the voice of the speaker so that it sounds like them, while the video is synthesised to synchronise the visual movement of the lips.
Prof Waibel said that this technique would be able to facilitate remote spoken communication in the future.
The technology could also potentially be used for other applications, including synthesising videos in different languages or for accurately lip-synching videos.
Prof Waibel has been working in the fields of speech recognition, processing and translation for decades, and the new technology builds on his prior work in these areas.
This includes a Lecture Translator app in use at KIT that can translate recorded speech into written text in real-time.
Today’s news was brought to you by TD SYNNEX – the UK’s number one solutions distributor.
Read more of our latest Industry Updates stories
