1 code implementation • 5 Feb 2024 • Amin Parchami-Araghi, Moritz Böhle, Sukrut Rao, Bernt Schiele
Knowledge Distillation (KD) has proven effective for compressing large teacher models into smaller student models.
1 code implementation • 19 Jun 2023 • Moritz Böhle, Navdeeppal Singh, Mario Fritz, Bernt Schiele
We present a new direction for increasing the interpretability of deep neural networks (DNNs) by promoting weight-input alignment during training.
1 code implementation • 23 Mar 2023 • Anna Kukleva, Moritz Böhle, Bernt Schiele, Hilde Kuehne, Christian Rupprecht
Such a schedule results in a constant `task switching' between an emphasis on instance discrimination and group-wise discrimination and thereby ensures that the model learns both group-wise features, as well as instance-specific details.
1 code implementation • 21 Mar 2023 • Sukrut Rao, Moritz Böhle, Amin Parchami-Araghi, Bernt Schiele
To gain a better understanding of which model-guiding approaches actually transfer to more challenging real-world datasets, in this work we conduct an in-depth evaluation across various loss functions, attribution methods, models, and 'guidance depths' on the PASCAL VOC 2007 and MS COCO 2014 datasets, and show that model guidance can sometimes even improve model performance.
1 code implementation • 21 Mar 2023 • Sukrut Rao, Moritz Böhle, Bernt Schiele
Finally, we propose a post-processing smoothing step that significantly improves the performance of some attribution methods, and discuss its applicability.
no code implementations • 20 Jan 2023 • Moritz Böhle, Mario Fritz, Bernt Schiele
Transformers increasingly dominate the machine learning landscape across many tasks and domains, which increases the importance for understanding their outputs.
1 code implementation • ICCV 2023 • Sukrut Rao, Moritz Böhle, Amin Parchami-Araghi, Bernt Schiele
To better understand the effectiveness of the various design choices that have been explored in the context of model guidance, in this work we conduct an in-depth evaluation across various loss functions, attribution methods, models, and 'guidance depths' on the PASCAL VOC 2007 and MS COCO 2014 datasets.
1 code implementation • CVPR 2022 • Sukrut Rao, Moritz Böhle, Bernt Schiele
Finally, we propose a post-processing smoothing step that significantly improves the performance of some attribution methods, and discuss its applicability.
1 code implementation • CVPR 2022 • Moritz Böhle, Mario Fritz, Bernt Schiele
We present a new direction for increasing the interpretability of deep neural networks (DNNs) by promoting weight-input alignment during training.
1 code implementation • 27 Sep 2021 • Moritz Böhle, Mario Fritz, Bernt Schiele
As a result, CoDA Nets model the classification prediction through a series of input-dependent linear transformations, allowing for linear decomposition of the output into individual input contributions.
1 code implementation • CVPR 2021 • Moritz Böhle, Mario Fritz, Bernt Schiele
Given the alignment of the DAUs, the resulting contribution maps align with discriminative input patterns.
1 code implementation • 18 Mar 2019 • Moritz Böhle, Fabian Eitel, Martin Weygandt, Kerstin Ritter
In this study, we propose using layer-wise relevance propagation (LRP) to visualize convolutional neural network decisions for AD based on MRI data.
2D Human Pose Estimation Quantitative Methods