The project delivers corpora of written and spoken language generated with the help of data augmentation for various domains relevant to SMEs and also provides easy-to-use tools which can be used to generate enough data from a small amount of user data via data augmentation to train voice assistants or other automatic speech processing systems.

Typical virtual teaching is not really interactive usually, as all content is recorded with static video and audio scripts and users can hardly provide feedback. Mixed Reality (MR), combined with animated volumetric avatars, makes immersed and interactive learning possible, taking the quality of teaching to a new level.

Aristech and the HNO clinic at the University Hospital Heidelberg are researching innovative systems ("Sprachversteher" project) to improve the understanding of people with impaired hearing due to old age.

Currently, text-to-speech synthesis sounds rather artificial, especially in emotional contexts, and merely offers limited emotionally expressive variations. As part of the Central Innovation Program for SMEs, Aristech worked on the EASY project together with the Technical University of Dresden. The goal was to narrate audio books with TTS voices which convey emotions realistically.

In large-scale damage situations, disasters and emergencies, emergency forces are faced with lots of information from different sources. In order to keep an up-to-date picture of the situation easily, this information has to be analyzed, processed and made available to specific users without distracting the emergency forces from their tasks.

The ZIM cooperation project's goal was to build a "virtual voice assistant for intelligent, context-sensitive speech dialogs". At the Innovation Day for SMEs 2019 on May 9 in Berlin, the project by the Heidelberg-based companies Aristech GmbH and sovanta AG was honored by the German Federal Ministry for Economic Affairs and Energy as the ZIM Cooperation Project of the Year.