Preshaping Hand Behaviour for Direct and Indirect Manipulation of 3D Objects
Authors
Paper Title
Preshaping Hand Behaviour for Direct and Indirect Manipulation of 3D Objects
Publication Info
- Topic area: Interaction techniques for 3D object manipulation in virtual environments.
- Keywords: Preshaping, 3D manipulation, direct interaction, indirect interaction, XR, virtual reality, controllers, gaze interaction, clutching, motor behaviour.
Background and Problem
- Problem / challenge: Current virtual interaction techniques alter natural hand-object interaction, affecting preparatory movements (preshaping) and necessitating compensatory strategies like clutching. The impact of interaction design on preshaping behaviour and its efficacy remains unclear.
- Significance: Understanding preshaping can improve interaction design, reducing clutching and enhancing performance in 3D manipulation tasks.
- Motivation and related work: Prior research has explored preshaping in physical and virtual contexts but has not systematically studied its role in direct versus indirect manipulation or its interaction with different input modalities (bare-hand vs. controller). This paper addresses these gaps.
Solution
- Proposed approach: A comparative study of preshaping behaviour in 3D docking tasks using four interaction techniques: VirtualHand, DirectController, Gaze&Pinch, and Gaze&Controller.
- Novelty:
- Analysis of how preshaping scales with task difficulty and varies across interaction techniques.
- Identification of trade-offs between motorspace constraints and sensory feedback in direct vs. indirect manipulation.
- Evidence linking preshaping deficits to increased clutching and compensatory strategies.
- Insights into how controllers mitigate preshaping deficits through in-hand rotation.
- Procedure and key techniques:
- Participants (N=20) performed a 6DOF docking task in VR under four conditions (direct/indirect, bare-hand/controller).
- Measured preshaping magnitude, effectiveness, preparation time, and manipulation metrics.
- Tasks involved translations and rotations (±45°, 90°, 135°) to evaluate preshaping responses.
Results
- Concrete findings:
- Direct techniques elicited more preshaping than indirect ones, particularly for smaller rotations.
- Bare-hand techniques induced more preshaping but were less effective than controllers for larger rotations.
- Indirect techniques resulted in reduced preshaping and longer manipulation times, compensated by more clutching.
- Controllers allowed in-hand rotation, extending range of motion and improving performance.
- Advantage over baselines:
- DirectController was the most effective technique, with the shortest task completion times and lowest clutch counts.
- Gaze&Pinch had the highest task load and lowest effectiveness due to reduced preshaping and reliance on compensatory strategies.
- Experiments / evaluation:
- Controlled 3D docking task with 5,703 valid trials.
- Metrics included preshaping magnitude, effectiveness, preparation/manipulation times, and subjective workload ratings (NASA-TLX).
- Statistical analysis revealed significant effects of directness, modality, and task rotation on performance.
- Limitations and future work:
- Study focused on abstract objects and specific interaction techniques; further research is needed on realistic applications and alternative modalities.
- Future work could explore preshaping in far-field interactions and develop anticipatory interfaces to guide users.
Summary
This study investigates preshaping behaviour in 3D object manipulation, revealing how interaction techniques (direct/indirect, bare-hand/controller) influence preparatory movements and task performance. Direct techniques and controllers generally supported more effective preshaping, reducing clutching and improving manipulation efficiency. Indirect and bare-hand techniques showed reduced preshaping, necessitating compensatory strategies. These findings provide actionable insights for designing XR systems that better support preshaping, enhance user performance, and reduce interaction overhead in virtual environments.
Research Questions / Practical Problems
Question signals indexed for this paper.
- 80%
Designing, Engineering, and Evaluating Gesture User Interfaces
CHI '18· Hand Gesture Recognition +1
- 80%
Towards a Consensus Gesture Set: A Survey of Mid-Air Gestures in HCI for Maximized Agreement Across Domains
CHI '23· Hand Gesture Recognition +1
- 80%
Spatial Hand Actions: Exploring the Hand Actions used to Represent Spatial Thinking for 3D Assembling Tasks
CHI '25· Hand Gesture Recognition +2
- 80%
Non-Natural Interaction Design
CHI '25· Hand Gesture Recognition +1
- 67%
TapGazer: Text Entry with Finger Tapping and Gaze-directed Word Selection
CHI '22· Hand Gesture Recognition +2
- 67%
Seeing and Touching the Air: Unraveling Eye-Hand Coordination in Mid-Air Gesture Typing for Mixed Reality
CHI '25· Hand Gesture Recognition +2
- 67%
GestureCanvas: A Programming by Demonstration System for Prototyping Compound Freehand Interaction in VR
UIST '23· Hand Gesture Recognition +2
Based on Jaccard similarity of research subtopics & professions (≥60%)