Parallel gesture recognition with soft real-time guarantees

Marr, Stefan; Renaux, Thierry; Hoste, Lode; Meuter, Wolfgang De

doi:10.1016/j.scico.2014.02.012

Cited by 3 publications

(3 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Existing centralized gesture recognition engines that recognize several hypothetical gestures simultaneously are making efforts to parallelize this processing while guaranteeing soft real-time (Marr et al 2014). In GestureAgents the processing of gestures in different applications would be done in parallel by definition, although without real-time guarantees.…”

Section: Discussionmentioning

confidence: 99%

Towards Concurrent Multi-Tasking in Shareable Interfaces

Juliá

Jordà

2015

Comput Supported Coop Work

View full text Add to dashboard Cite

Shareable interfaces, those that can be interacted simultaneously by several users, are a common tool used both in CSCW research and in real world applications. They tend however to lack a capability that has been traditionally relevant to the usefulness of computing systems: multi-tasking. In this paper we explain why a combination of the multiuser features of shareable interfaces and the multitasking capabilities of general-purpose computing, could be relevant for building useful systems, and why these features are not present today in most of the current prototypes and systems. We also discuss possible approaches for solving the problems that prevent shareable interfaces to fully support multitasking, and we present a novel approach based on a distributed, application-centered, content-based gesture disambiguation. We describe how an already existing framework, GestureAgents, implements this new approach, focusing on expanding the description of the relevant elements related to this problem, and conclude with some example applications and a discussion.

show abstract

Section: Discussionmentioning

confidence: 99%

Towards Concurrent Multi-Tasking in Shareable Interfaces

Juliá

Jordà

2015

Comput Supported Coop Work

View full text Add to dashboard Cite

show abstract

“…The declarative nature of Midas, coupled with the modular approach of Mudra has allowed individual parts of the architecture to be replaced, e. g., to increase performance by enabling the rule engine to employ 4 [13], or to distribute workload over multiple machines [14].…”

Section: Validationmentioning

confidence: 99%

Software Engineering Principles in the Midas Gesture Specification Language

Renaux

Hoste

Scholliers

et al. 2014

Proceedings of the 2nd Workshop on Programming for Mobile &Amp; Touch - PROMOTO '14

Self Cite

View full text Add to dashboard Cite

While many technologies for gesture-based interaction have been proposed and implemented, few focus on core software engineering principles that are commonplace in traditional programming languages. The lack of such principles restricts the applicability of those technologies when developing large scale gesture enabled systems. This paper describes the software engineering challenges associated with developing multi-touch gesture-based interaction, and proposes a solution in the form of the Midas declarative gesture specification language. Midas embeds concepts of logical programming languages and complex event processing to ease the development of gesture based applications. We show how it can be applied to multi-touch gesture recognition, and evaluated our solution in real-world applications.

show abstract

“…Thread and data parallelization can both be applied for further improving the performance of the inference process. Existing works [23,2] deal with parallelizing the core Rete algorithm. In addition to parallelizing the core fact processing step with multiple worker threads during the fact processing step, we study where and how data parellelism -namely vectorization -can be applied.…”

Section: Introductionmentioning

confidence: 99%

Hiperfact: In-Memory High Performance Fact Processing -- Rethinking the Rete Inference Algorithm

Indiono,

Rinderle-Ma

2020

Preprint

View full text Add to dashboard Cite

The Rete forward inference algorithm forms the basis for many rule engines deployed today, but it exhibits the following problems:(1) the caching of all intermediate join results, (2) the processing of all rules regardless of the necessity to do so (stemming from the underlying forward inference approach), (3) not defining the join order of rules and its conditions, significantly affecting the final run-time performance, and finally (4) pointer chasing due to the overall network structure, leading to inefficient usage of the CPU caches caused by random access patterns. The Hiperfact approach aims to overcome these shortcomings by (1) choosing cache efficient data structures on the primary rank 1 fact index storage and intermediate join result storage levels, (2) introducing island fact processing for determining the join order by ensuring minimal intermediate join result construction, and (3) introducing derivation trees to allow for parallel read/write access and lazy rule evaluation. The experimental evaluations show that the Hiperfact prototype engine implementing the approach achieves significant improvements in respect to both inference and query performance. Moreover, the proposed Hiperfact engine is compared to existing engines in the context of a comprehensive benchmark.

show abstract

Parallel gesture recognition with soft real-time guarantees

Cited by 3 publications

References 25 publications

Towards Concurrent Multi-Tasking in Shareable Interfaces

Towards Concurrent Multi-Tasking in Shareable Interfaces

Software Engineering Principles in the Midas Gesture Specification Language

Hiperfact: In-Memory High Performance Fact Processing -- Rethinking the Rete Inference Algorithm

Contact Info

Product

Resources

About