Fully articulated hand tracking promises to enable fundamentally new interactions with virtual and augmented worlds, but the limited accuracy and efficiency of current systems has prevented widespread adoption. Today's dominant paradigm uses machine learning for initialization and recovery followed by iterative model-fitting optimization to achieve a detailed pose fit. We follow this paradigm, but make several changes to the model-fitting, namely using: (1) a more discriminative objective function; (2) a smooth-surface model that provides gradients for non-linear optimization; and (3) joint optimization over both the model pose and the correspondences between observed data points and the model surface. While each of these changes may actually increase the cost per fitting iteration, we find a compensating decrease in the number of iterations. Further, the wide basin of convergence means that fewer starting points are needed for successful model fitting. Our system runs in real-time on CPU only, which frees up the commonly over-burdened GPU for experience designers. The hand tracker is efficient enough to run on low-power devices such as tablets. We can track up to several meters from the camera to provide a large working volume for interaction, even using the noisy data from current-generation depth cameras. Quantitative assessments on standard datasets show that the new approach exceeds the state of the art in accuracy. Qualitative results take the form of live recordings of a range of interactive experiences enabled by this new approach.
We present a machine learning technique for estimating absolute, per-pixel depth using any conventional monocular 2D camera, with minor hardware modifications. Our approach targets close-range human capture and interaction where dense 3D estimation of hands and faces is desired. We use hybrid classification-regression forests to learn how to map from near infrared intensity images to absolute , metric depth in real-time. We demonstrate a variety of human-computer interaction and capture scenarios. Experiments show an accuracy that outperforms a conventional light fall-off baseline, and is comparable to high-quality consumer depth cameras, but with a dramatically reduced cost, power consumption, and form-factor.
We present findings from a year-long engagement with a street and its community. The work explores how the production and use of data is bound up with place, both in terms of physical and social geography. We detail three strands of the project. First, we consider how residents have sought to curate existing data about the street in the form of an archive with physical and digital components. Second, we report endeavours to capture data about the street's environment, especially of vehicle traffic. Third, we draw on the possibilities afforded by technologies for polling opinion. We reflect on how these engagements have: materialised distinctive relations between the community and their data; surfaced flows and contours of data, and spatial, temporal and social boundaries; and enacted a multiplicity of 'small worlds'. We consider how such a conceptualisation of data-in-place is relevant to the design of technology.
Considerable resources have been spent developing and rigorously testing HIV prevention intervention models, but such models do not impact the AIDS pandemic unless they are implemented effectively by community-based organizations (CBOs) and health departments. The Mpowerment Project (MP) is being implemented by CBOs around the U.S. It is a multilevel, evidence-based HIV prevention program for young gay/bisexual men that targets individual, interpersonal, social, and structural issues by using empowerment and community mobilization methods. This paper discusses the development of an intervention to help CBOs implement the MP called the Mpowerment Project Technology Exchange System (MPTES); CBOs' uptake, utilization and perceptions of the MPTES components; and issues that arose during technical assistance. The seven-component MPTES was provided to 49 CBOs implementing the MP that were followed longitudinally for up to two years. Except for the widely used program manual, other program materials were used early in implementing the MP and then their use declined. In contrast, once technical assistance was proactively provided, its usage remained constant over time, as did requests for technical assistance. CBOs expressed substantial positive feedback about the MPTES, but felt that it needs more focus on diversity issues, describing real world implementation approaches, and providing guidance on how to adapt the MP to diverse populations.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.