WaveSense to Use Geo Radars For Autonomous Vehicles

Company believes that this can improve the safety of self-driving cars
22 August 2018   415

WaveSense executive director Tarik Bolat announced that radars, lidars and cameras do not provide full-fledged safety of self-driving cars. In rain, snow and fog, sensors may not recognize road signs or markings and make mistakes when constructing a route. To solve this problem, he proposed to work the way through the maps of underground objects.

It is proposed to use georadars for orientation on the underground map. They are equipped with 12-element antenna arrays to transmit high-frequency magnetic pulses. Waves penetrate to a depth of three meters below the earth's surface and are reflected from underground objects: pipes, stones and tree roots. The received signals are checked against the underground WaveSense card and help the autopilot system to determine the location of the car.

The technology was developed at the Lincoln Laboratory at the Massachusetts Institute of Technology. It was intended for military equipment used in regions with bad roads without road signs.

The effectiveness of the system was proved in 2016. Georadar was installed on a sports truck, and in a snowstorm the car remained on its lane with an error of not more than 4 cm.

The head of WaveSense admitted that the technology is not capable of completely replacing radars, lidars and cameras. It is proposed to be used as an addition to existing sensors operating under conditions of poor visibility. With all its advantages, the WaveSense system is difficult to deploy, as each road needs individual scanning.

Microsoft to Use AI to Create Human Voice

Synthetic voice is nearly indistinguishable from recordings of people
27 September 2018   457

Researchers from Microsoft recorded computer voice, imitating human speech. To overcome the difficulties of the traditional model, they used neural networks for speech synthesis. Microsoft promises to provide support for 49 languages ​​and the ability to create unique voices for the needs of companies in the near future.

Synthesis of speech with the help of neural networks involves comparing the stress and length (so-called prosody) of the speaker's speech units, as well as their synthesis into a computer voice. In systems of traditional speech synthesis, prosody is divided into acoustic and linguistic analysis, controlled by various models. As a result, the speech is noisy and indistinct. Representatives of Microsoft argue that in the model of neural synthesis two stages are combined into one, so the voice sounds like a real one.

The developers are convinced that the synthesis of speech with the help of neural networks will make it more natural to communicate with virtual interlocutors and assistants. Moreover, it will enable you to convert e-books into audiobooks and will allow you to change the scoring of built-in navigators.

Microsoft Neural TTS
Microsoft Neural TTS

Azure computing power is available for real-time use, and Azure Kubernetes is responsible for this. Simultaneous application of neural synthesis of speech together with traditional speaks about expansion and increase of availability of service. At the moment, there are a female voice named Jessa and a man named Guy.

Microsoft is competing in speech recognition and synthesis technologies with Google, which updated its services in late August 2018. Google Cloud announced the release of a stable API for the synthesis of speech Cloud Text-to-Speech with the experimental function of audio profiles and support for several new languages.