On April 28th, I had the opportunity to visit the Computer Vision and Multimodal Computing Department at the Max Planck Institute for Informatics which is headed by Prof. Bernt Schiele. During my visit I also had the opportunity to talk to Björn Andres, Mario Fritz and Andreas Bulling — among others. Aside from getting to know their latest research, I also had great conversations about research in general and PhD programs in specific.
In my talk, I presented results from my bachelor thesis and my recent CVIU publication. In the former, I tried to improve superpixel algorithms using depth information. Afterwards, I worked on an extensive comparison of 28 different superpixel algorithms on 5 different datasets. The main contribution is a final ranking including clear recommendations which superpixel algorithms to use in practice. Interactive visualizations of the results can also be found here: Qualitative Results; Quantitative Results — as shown in Figure 1.
The presentation slides can be downloaded below:
The presentation includes GIF animations — they have been tested with Adobe Reader and Foxit Reader on Windows. In Chrome and Firefox, the animations might not work.
Talk SlidesPersonal Take-Away. As a master student I can only recommend to take any chance to visit other universities or departments. While I also discussed amazing research in different directions — including directions I did not know so far —, I benefited most from the informal discussions about career in general (in academia and industry).
What is your opinion on this article? Did you find it interesting or useful? Let me know your thoughts in the comments below: