Our »Audio and Media Technologies« division develops world-class audio processing technology. It is one of the largest organizations in the world for audio and speech processing. Over 200 engineers and scientists develop first-rate audio technology which is sold world-wide, in Europe, USA, China, Korea and Japan.
In the past 25 years, Fraunhofer IIS repeatedly won international competitions in the audio and media field: e.g. with mp3 and codecs of the AAC-family. You will be part of the audio and media team.
A conversation between two humans can be seen as a chain of statements, which are somehow connected to each other. In a conversation between a human and a machine (let’s say a chat bot), the same model can be applied. Conventionally the machine waits until the human is done
talking / typing and interprets the whole message. This is different from the human conversation. The “listening” human constructs continuously a mental representation of what he or she understood up till then. In addition, there can also be signs of understanding or misunderstanding to influence the “talking” human and therefore the upcoming message.
The goal of this work, is to create and benchmark a machine learning based system, which implements this “human-like” behavior. This includes addressing a problem, which relies on Natural Language Understanding, combining it with an incremental learning component and evaluating the quality of the systems. The most interesting result of this project should be the differences in interactions between “human-like” and “non-human-like” systems.
Your responsibilities: You…
- do research of relevant literature
- are implementing of a model for an NLU task
- are involved in the development of an incremental learning method and its integration
- do performance evaluation in the developed method regarding User Experience and scalability
Was du mitbringst
- You are familiar with the Python program language
- You have experience with machine learning methods
Was du erwarten kannst
- An open and cooperative working environment
- Collaboration in interesting and innovative projects
- Many opportunities to gain practical experience and attend seminars
- Flexibility concerning your working hours
The thesis will be assigned and carried out in accordance with the rules of your university. For this reason, please discuss the thesis with a professor who can advise you over the course of the project.
Applications are possible in German and English. Please include a cover letter, your CV and your latest transcripts of records (as PDF) and quote ID number 50224. Address your application to Nina Wörlein.
Please let us know how you learned about this job opportunity.