On 18.01.19 Prof. Alexander Waibel visited MSRM and gave a detailed talk about "Intelligence Systems and a Language Transparent World". Prof. Waibel is one the world’s leading experts in speech recognition and translation as well as human communication signals and systems.
After centuries of separation and misunderstandings, we are lucky to be living in the generation that can see an end to language/cultural barriers between the peoples of our planet. AI, specifically: automatic recognition, translation and understanding of language in all its forms (text, speech, images, tactile, body, emotion,…) is now becoming possible and ubiquitous on the internet, making it feasible for all of us to operate and communicate in any country, language, mode and situation. The key to these advances has been machine learning combined with an exponential growth in computing and language data. After dismissing it as impossible just a few decades ago, we are now seeing speech and language systems that approach/exceed human performance. Is the problem solved? Will it replace human translators, interpreters, and other knowledge workers? In our view, a symbiotic relationship between humans and machines is more likely and more desirable. In this spirit, our systems are designed to support and augment human communication and abilities.
In my talk, I will discuss systems we have designed and deployed to overcome language and communication barriers, including:
- Interpreting Smartphones for tourists, healthcare and relief workers
- Communicators on tablets in Humanitarian and Government Missions
- Road sign interpreters that translate road signs while traveling abroad
- Multimodal interpretation of human meta-communication (emotion, gaze..)
- Multilingual subtitling and translation of TV broadcasts
- Automatic Interpretation of lectures for foreign students
- Symbiotic human & machine interpreters at the European Parliament
- Translation of silent speech I will review the technology, algorithmic advances, progress in performance and usability, and discuss remaining scientific challenges.
Dr. Alexander Waibel is Professor of Computer Science at Carnegie Mellon University (USA) and at the Karlsruhe Institute of Technology (Germany). He is director of the International Center for Advanced Communication Technologies. Waibel is internationally known for his work on AI, Machine Learning, Multimodal Interfaces and Speech Translation Systems. He and his team developed the TDNN, the first shift-invariant “Convolutional” Neural Network, early multimodal interfaces and the first speech translation and simultaneous interpreting systems. Waibel is a member of the National Academy of Sciences of Germany and a Fellow of the IEEE. He received BS, MS and PhD degrees from MIT and CMU, respectively.