Human-in-the-loop Video Mining

This project took place in summer term 2021, you CAN NOT apply to this project anymore!

Results of this project are explained in detail in the final documentation .

Design AI GmbH is a Munich based AI start-up supported by the TUM, UnternehmerTUM, and NVIDIA. We develop cutting edge machine learning / deep learning solutions by combining Design Thinking and Artificial Intelligence.

The goal of this project is to further improve our complex model architecture in order to set new state-of-the-art performances of our deep learning and improve the overall user experience. Possible directions are modification of the feature extraction subnetworks, improvement of the core network, addition of more modalities, improvement of the continual / active learning framework, integration of data augmentation, ensemble methods or uncertainty estimations, and any other improvement that provides customer value.

We will provide a dataset with over 20.000 labeled videos, a working prototype and 10.000 $ worth of cloud resources for further large-scale experimentation. If interested, a follow-up via a master thesis or working student position is possible after successful completion of this project.

Project Tasks

  • Identification of problems and improvement opportunities for our product

  • Translation of those insights into technical requirements for our current architecture

  • Research of state-of-the-art approaches to fulfill those requirements

  • Development and testing of new or improved architectures / modules

Your Profile

  • Strong foundations in ML and DL •

  • First DL research experiences, e.g. via Advanced DL courses at TUM

  • Fluency in PyTorch

  • Basic understanding of Continual / Online / Active Learning

  • Interest to work with agile software development in a fast-paced start-up

Why Design AI

  • Learn how to build a globally innovative AI product based on state-of-the-art technology

  • Experiment as much as you want – up to 10.000 $ worth of AWS credits for your free usage

  • Work in an growing, TUM-start-up and get close supervision by co-founders and DI-Lab alumni

Questions? Write us: