Franz Müller

Arman Kazemi, Mohammad Mehdi Sharifi, Ann Franchesca Laguna et al.

Nearest neighbor (NN) search is an essential operation in many applications, such as one/few-shot learning and image classification. As such, fast and low-energy hardware support for accurate NN search is highly desirable. Ternary content-addressable memories (TCAMs) have been proposed to accelerate NN search for few-shot learning tasks by implementing $L_\infty$ and Hamming distance metrics, but they cannot achieve software-comparable accuracies. This paper proposes a novel distance function that can be natively evaluated with multi-bit content-addressable memories (MCAMs) based on ferroelectric FETs (FeFETs) to perform a single-step, in-memory NN search. Moreover, this approach achieves accuracies comparable to floating-point precision implementations in software for NN classification and one/few-shot learning tasks. As an example, the proposed method achieves a 98.34% accuracy for a 5-way, 5-shot classification task for the Omniglot dataset (only 0.8% lower than software-based implementations) with a 3-bit MCAM. This represents a 13% accuracy improvement over state-of-the-art TCAM-based implementations at iso-energy and iso-delay. The presented distance function is resilient to the effects of FeFET device-to-device variations. Furthermore, this work experimentally demonstrates a 2-bit implementation of FeFET MCAM using AND arrays from GLOBALFOUNDRIES to further validate proof of concept.

Denys J. C. Matthies, Felix M. Manke, Franz Müller et al.

Compared to real world tasks, completing tasks in a virtual environment (VE) seldom involves the whole spectrum of skills the human body offers. User input in a VE is commonly accomplished through simple finger gestures, such as walking in a scene by simply pressing a button, even if this kind of interaction is not very suitable. In order to create a more intuitive and natural interaction, diverse projects try to tackle the problem of locomotion in VEs by trying to enable a natural walking movement, which is also supposed to increase the level of immersion. Existing solutions such as treadmills are still expensive and need additional fixation of the body. In this paper, we describe a simple and inexpensive way to build a useful locomotion interface using a conventional sports stepper and an Arduino. This device enables control in a VE by walking-in-place and without the need for any additional fixation gadgets. We conducted a user study with 10 participants to evaluate the impression on the joy and ease of use, immersion and reliability in comparison to other interfaces used for locomotion, such as the Wii Balance Board and a Wand Joystick. We found out that the stepper is experienced slightly better in terms of immersion and joy of use. Furthermore, found that pressing buttons on a Joystick was perceived to be more reliable.