Acquisition of off-screen object by predictive jumping

Kazuki Takashima, Sriram Subramanian, Takayuki Tsukitani, Yoshifumi Kitamura, Fumio Kishino

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)


We propose predictive jumping (PJ), a fast and efficient algorithm that enables user navigation to off-screen targets. The algorithm is inspired by Delphian Desktop [1] and the off-screen visualization technique-Halo [2]. The Halos represented at the edge of the viewport help users estimate off-screen target distance and encourage them to make a single fluid mouse movement toward the target. Halfway through the user's motion, the system predicts the user's intended target and quickly moves the cursor towards that predicted off-screen location. In a pilot study we examine the user's ability to select off-screen targets with predictive models based on user's pointing kinematics for off-screen pointing with Halo. We establish a linear relationship between peak velocity and target distance for PJ. We then conducted a controlled experiment to evaluate PJ against other Halo-based techniques, Hop [8] and Pan with Halo. The results of the study highlight the effectiveness of PJ.

Original languageEnglish
Title of host publicationComputer-Human Interaction - 8th Asia-Pacific Conference, APCHI 2008, Proceedings
Number of pages10
Publication statusPublished - 2008
Event8th Asia-Pacific Conference on Computer-Human Interaction, APCHI 2008 - Seoul, Korea, Republic of
Duration: 2008 Jul 62008 Jul 9

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5068 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference8th Asia-Pacific Conference on Computer-Human Interaction, APCHI 2008
Country/TerritoryKorea, Republic of


  • Kinematics
  • Mouse
  • Prediction
  • Scroll interaction technique


Dive into the research topics of 'Acquisition of off-screen object by predictive jumping'. Together they form a unique fingerprint.

Cite this