TOBIAS HINZ
PhD Candidate
University of Hamburg
I am a 4th-year PhD student at the Knowledge Technology group at the University of Hamburg, advised by Professor Wermter. I work on generative models that learn visual representation in a semi- or unsupervised way. This includes applications such as unconditional image generation, text-to-image synthesis, and disentangled representations. I am also interested in the automatic evaluation of generative models in the visual domain, which is a challenging problem.
I completed my Bachelor degree in Business Informatics at the University of Mannheim, during which I also studied at the National University of Singapore for one semester. Following that, I did a research oriented Master degree in Intelligent Adaptive Systems at the University of Hamburg.
News
- 06/2020: I started my research internship with Adobe. Big thanks to Adobe for facilitating internships via home-office.
- 03/2020: We posted our work on Single-Image GANs with the available code and blog post.
- 01/2020: I will join Matthew Fisher and Oliver Wang at Adobe in San Francisco this summer for a research internship.
- 10/2019: We uploaded our latest work on generative text-to-image synthesis. Check out our blog post and implementation.
- 08/2019: I attended the DLRLSS in Edmonton. Have a look at the new blog post for some highlights.
- 07/2019: I attended this year's EEML in Bucharest and summarized my experiences in a new blog post.
- 05/2019: I just returned from this year's ICLR. Check out my blog post and see this post which mentions our paper.
- 01/2019: I started a new position with the Crossmodal Learning (CML) project (subproject A5) at the University of Hamburg.
- 12/2018: Our work on controlled image generation got accepted to ICLR 2019.
- 07/2018: I attended this year's IEEE World Congress on Computational Intelligence (WCCI) in Rio de Janeiro (Brazil).
Improved Techniques for Training Single-Image GANs
Posted on 24 Mar 2020 | Reading time: 7 minutes
Overview of our paper about training GANs on a single image for tasks such as image generation, image harmonization, and image animation.
Semantic Object Accuracy for Generative Text-to-Image Synthesis
Posted on 30 Oct 2019 | Reading time: 8 minutes
Overview of our paper about our new model and evaluation metric for generative text-to-image synthesis models.
My Time At The Deep Learning And Reinforcement Learning Summer School (DLRLSS) in Edmonton (Canada)
Posted on 06 Aug 2019 | Reading time: 17 minutes
After attending the EEMLSS I was lucky enough to also attend the Deep Learning And Reinforcement Learning Summer School (DLRLSS) in Edmonton (Canada) 2019.
Impressions from the Eastern European Machine Learning Summer School (EEML) 2019 in Bucharest (Romania)
Posted on 08 Jul 2019 | Reading time: 6 minutes
My experiences at the Eastern European Machine Learning Summer School (EEML) organized by DeepMind in Bucharest (Romania) 2019.
See All
Selected Publications
For a full list, have a look at my Google Scholar page.
Speeding Up the Hyperparameter Optimization Of Deep Convolutional Neural Networks
How to use lower dimensional data representations to speed up the hyperparameter optimization for CNNs processing images..
T. Hinz, N. Navarro-Guerrero, S. Magg, S. Wermter, International Journal of Computational Intelligence and Applications 2018.
See All