Fabian Groh

• University of Tübingen

Approximate nearest neighbor (ANN) search in high dimensions is an integral part of several computer vision systems and gains importance in deep learning with explicit memory representations. Since PQT (Wieschollek et al. , 2016), FAISS (Johnson et al. , 2021), and SONG (Zhao et al. , 2020) started to leverage the massive parallelism offered...

Approximate nearest neighbor (ANN) search in high dimensions is an integral part of several computer vision systems and gains importance in deep learning with explicit memory representations. Since PQT and FAISS started to leverage the massive parallelism offered by GPUs, GPU-based implementations are a crucial resource for today's state-of-the-art...

Traditional convolution layers are specifically designed to exploit the natural data representation of images – a fixed and regular grid. However, unstructured data like 3D point clouds containing irregular neighborhoods constantly breaks the grid-based data assumption. Therefore applying best-practices and design choices from 2D-image learning met...

The goal of this work is to enable deep neural networks to learn representations for irregular 3D structures -- just like in common approaches for 2D images. Unfortunately, current network primitives such as convolution layers are specifically designed to exploit the natural data representation of images -- a fixed and regular grid structure. This...

We introduce a highly accurate and precise multi-view, multi-projector, and multi-pattern phase scanning method for shape acquisition that is able to handle occlusions and optically challenging materials. The 3D reconstruction is formulated as a two-step process which first estimates reliable measurement samples and then simultaneously optimizes ov...

Fisher-Vectors (FV) encode higher-order statistics of a set of multiple local descriptors like SIFT features. They already show good performance in combination with shallow learning architectures on visual recognitions tasks. Current methods using FV as a feature descriptor in deep architectures assume that all original input features are static. W...

The Multifunctional and Adaptive Meal Preparation Facility (MAMPF) is a versatile and adaptive kitchen system allowing even an inexperienced user to create ambitious dishes. On the hardware side, the system controls hotplates by regulating the temperature of heated regions to turn the attention to the food actually being heated instead of the hotpl...