An official website of the United States government
Abstract Objective.The RSVP Keyboard is a non-implantable, event-related potential-based brain-computer interface (BCI) system designed to support communication access for people with severe speech and physical impairments. Here we introduce inquiry preview (IP), a new RSVP Keyboard interface incorporating switch input for users with some voluntary motor function, and describe its effects on typing performance and other outcomes.Approach.Four individuals with disabilities participated in the collaborative design of possible switch input applications for the RSVP Keyboard, leading to the development of IP and a method of fusing switch input with language model and electroencephalography (EEG) evidence for typing. Twenty-four participants without disabilities and one potential end user with incomplete locked-in syndrome took part in two experiments investigating the effects of IP and two modes of switch input on typing accuracy and speed during a copy-spelling task.Main results.For participants without disabilities, IP and switch input tended to worsen typing performance compared to the standard RSVP Keyboard condition, with more consistent effects across participants for speed than for accuracy. However, there was considerable variability, with some participants demonstrating improved typing performance and better user experience (UX) with IP and switch input. Typing performance for the potential end user was comparable to that of participants without disabilities. He typed most quickly and accurately with IP and switch input and gave favorable UX ratings to those conditions, but preferred standard RSVP Keyboard.Significance.IP is a novel multimodal interface for the RSVP Keyboard BCI, incorporating switch input as an additional control signal. Typing performance and UX and preference varied widely across participants, reinforcing the need for flexible, customizable BCI systems that can adapt to individual users. ClinicalTrials.gov Identifier: NCT04468919.
more »
« less
Meta Design Studies: A Structured Approach for Deriving Domain-Oriented Visualization Recommendation Strategies
We introduce the concept of a meta design study as a structured approach to extract information from design study papers for the development of generalized tools in specific problem areas or domains. We explore the potential of meta design studies for creating domain-oriented visualization recommendation (VisRec) strategies. To demonstrate this concept, we present RSVP, a system derived from a meta design study conducted on Visual Parameter Space Analysis (VPSA). We outline the individual steps of the meta design study, highlight key concepts of the resulting VisRec strategy, and present a non-obtrusive implementation of this approach in RSVP.
more »
« less
- Award ID(s):
- 2007436
- PAR ID:
- 10465557
- Date Published:
- Journal Name:
- IEEE Visualization Conference Poster Proceedings
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract Metasurfaces are planar structures that can offer unprecedented freedoms to manipulate electromagnetic wavefronts at deep‐subwavelength scale. The wavelength‐dependent behavior of the metasurface could severely reduce the design freedom. Besides, realizing high‐efficiency metasurfaces with a simple design procedure and easy fabrication is of great interest. Here, a novel approach to design highly efficient meta‐atoms that can achieve full 2π phase coverage at two wavelengths independently in the transmission mode is proposed. More specifically, a bilayer meta‐atom is designed to operate at two wavelengths, the cross‐polarized transmission efficiencies of which reach more than 70% at both wavelengths. The 2π phase modulations at two wavelengths under the circularly polarized incidence can be achieved independently by varying the orientations of the two resonators constructing the meta‐atom based on Pancharatnam–Berry phase principle. As proof‐of‐concept demonstrations, three dual‐wavelength meta‐devices employing the proposed meta‐atom are numerically investigated and experimentally verified, including two metalenses (1D and 2D) with the same focusing length and a vortex beam generator carrying different orbital angular momentum modes at two operation wavelengths. Both the simulation and experimental results satisfy the design goals, which validate the proposed approach.more » « less
-
Abstract Data driven advanced manufacturing research is dependent on access to large datasets made available from across the product lifecycle — from the concept design phase all the way down to end use and disposal. Despite such data being generated at a rapid pace, most product design data is archived in inaccessible silos. This is particularly acute in academic research laboratories and with data generated during product design and manufacturing courses. This project seeks to create an infrastructure that allow users (academia and the general public) to easily upload project data and related meta-data. Current manufacturing research must shift from siloed repositories of product manufacturing data to a federated, decentralized, open and inter-operable approach. In this regard, we build ‘FabWave’ a cyber-infrastructure tool designed to capture manufacturing data. In its first pilot implementation, we focused our attention to gathering information rich 3D Mechanical CAD data and related meta-data associated with them, with the intent to make it easier for users to upload and access product design data. We describe workflows that we have initially tested out within the two academic universities and under two different course structures. We have also developed automated workflows to gather license appropriate CAD assemblies from commercial repositories. Our intent is to create the only known largest available CAD model set within academia for enabling research in data-driven computational research in digital design, fabrication and quality control.more » « less
-
Learning-to-learn (using optimization algorithms to learn a new optimizer) has successfully trained efficient optimizers in practice. This approach relies on meta-gradient descent on a meta-objective based on the trajectory that the optimizer generates. However, there were few theoretical guarantees on how to avoid meta-gradient explosion/vanishing problems, or how to train an optimizer with good generalization performance. In this paper, we study the learning-to-learn approach on a simple problem of tuning the step size for quadratic loss. Our results show that although there is a way to design the meta-objective so that the meta-gradient remain polynomially bounded, computing the meta-gradient directly using backpropagation leads to numerical issues that look similar to gradient explosion/vanishing problems. We also characterize when it is necessary to compute the meta-objective on a separate validation set instead of the original training set. Finally, we verify our results empirically and show that a similar phenomenon appears even for more complicated learned optimizers parametrized by neural networks.more » « less
-
null (Ed.)Learning-to-learn – using optimization algorithms to learn a new optimizer – has successfully trained efficient optimizers in practice. This approach relies on meta-gradient descent on a meta-objective based on the trajectory that the optimizer generates. However, there were few theoretical guarantees on how to avoid meta-gradient explosion/vanishing problems, or how to train an optimizer with good generalization performance. In this paper we study the learning-to-learn approach on a simple problem of tuning the step size for quadratic loss. Our results show that although there is a way to design the meta-objective so that the meta-gradient remain polynomially bounded, computing the meta-gradient directly using backpropagation leads to numerical issues that look similar to gradient explosion/vanishing problems. We also characterize when it is necessary to compute the meta-objective on a separate validation set instead of the original training set. Finally, we verify our results empirically and show that a similar phenomenon appears even for more complicated learned optimizers parametrized by neural networks.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
