no code implementations • 14 Mar 2021 • Manoj Rohit Vemparala, Alexander Frickenstein, Nael Fasfous, Lukas Frickenstein, Qi Zhao, Sabine Kuhn, Daniel Ehrhardt, Yuankai Wu, Christian Unger, Naveen Shankar Nagaraja, Walter Stechele
The distilled models exhibit their strength against all white box attacks with an exception of C&W.
no code implementations • 7 Jan 2021 • Manoj-Rohit Vemparala, Nael Fasfous, Alexander Frickenstein, Mhd Ali Moraly, Aquib Jamal, Lukas Frickenstein, Christian Unger, Naveen-Shankar Nagaraja, Walter Stechele
In this context, we present Learning to Prune Faster which details a multi-task, try-and-learn method, discretely learning redundant filters of the CNN and a continuous action of how long the layers have to be fine-tuned.
no code implementations • 3 Nov 2020 • René Schuster, Christian Unger, Didier Stricker
Motion estimation is one of the core challenges in computer vision.
no code implementations • 21 Oct 2020 • René Schuster, Christian Unger, Didier Stricker
Contrary to the ongoing trend in automotive applications towards usage of more diverse and more sensors, this work tries to solve the complex scene flow problem under a monocular camera setup, i. e. using a single sensor.
no code implementations • 21 Aug 2020 • René Schuster, Oliver Wasenmüller, Christian Unger, Didier Stricker
Interpolation of sparse pixel information towards a dense target resolution finds its application across multiple disciplines in computer vision.
no code implementations • 27 Jul 2020 • Alexander Frickenstein, Manoj-Rohit Vemparala, Nael Fasfous, Laura Hauenschild, Naveen-Shankar Nagaraja, Christian Unger, Walter Stechele
Model compression techniques, such as pruning, are emphasized among other optimization methods for solving this problem.
no code implementations • 15 Jun 2020 • Alexander Frickenstein, Manoj Rohit Vemparala, Jakob Mayr, Naveen Shankar Nagaraja, Christian Unger, Federico Tombari, Walter Stechele
The driveable area detection, posed as a two class segmentation task, can be efficiently modeled with slim binary networks.
no code implementations • 12 Apr 2019 • René Schuster, Oliver Wasenmüller, Christian Unger, Didier Stricker
Not only the tuning of hyperparameters, but also the gathering and selection of training data, the design of the loss function, and the construction of training schedules is important to get the most out of a model.
no code implementations • 5 Apr 2019 • René Schuster, Oliver Wasenmüller, Christian Unger, Didier Stricker
Our network has a very large receptive field and avoids striding layers to maintain spatial resolution.
no code implementations • 26 Feb 2019 • René Schuster, Oliver Wasenmüller, Christian Unger, Georg Kuschk, Didier Stricker
State-of-the-art scene flow algorithms pursue the conflicting targets of accuracy, run time, and robustness.