ADAPTIVE BUBBLE VR / SPATIAL INTERACTION

Intent
in motion.

Find the target.
Feel the difference.

Motion and density guide VR target selection.Unity / OpenXR · Adaptive intent scoring.

Try the idea ↓Code ↗

Actual Unity VR prototype with crowded yellow objects, a green recommended target, score and timing display — Green marks the current recommendation.
Original prototype screenshot · presentation slide 4.

PROXIMITY+MOTION+LOCAL DENSITY

Three signals. One recommendation.

01 / EXPLORE THE MECHANISM

Point toward
what you mean.

Move through the field and watch the recommendation change. Compare distance-only scoring with the weighted, density-aware intent score used by the prototype.

2D SCORING ILLUSTRATION

MOVE POINTER OR DRAG · ARROW KEYS SUPPORTED

12-candidate 2D scoring illustration. Move the controller to explore distance, motion, and local density.

02 / HOW IT WORKS

Adapt the weights.
Reveal the intent.

WDIB stands for Weighted and Density-Aware Intent Bubble. In dense regions, proximity receives more weight; in sparse regions, the direction of motion contributes more.

01 / DISTANCE

How close is it?

A smooth, clamped distance score favors nearby targets inside the configured detection range.

Dᵢ = clamp(1 − dᵢ / R, 0, 1)

02 / TRAJECTORY

Where are you heading?

The alignment between hand movement and the direction to each candidate estimates the next intended target.

Tᵢ = max(0, v̂ · p̂ᵢ)

03 / DENSITY

How crowded is it?

The nearby candidate count shifts the balance between distance and directional alignment.

ρ = clamp(Nnear / 12, 0, 1)

THE RELEASED CODE

Sᵢ = (0.7 + 0.4ρ) Dᵢ + 0.3(1 − ρ) Tᵢ

Exposed script defaults. Unity Inspector settings can override them.

03 / THREE WAYS TO SELECT

From contact
to recommendation.

The study compared direct hand collision, a fixed bubble, and WDIB within the same participant group.

Original hand-selection condition with a crowded field of yellow targets — **01 / Hand collision**
Direct hand movement and contact.

Original fixed-bubble selection condition with a red highlighted candidate — **02 / Fixed bubble**
Distance-driven feedback with a red highlight.

Original WDIB condition with green target recommendation — **03 / WDIB**
Distance and motion inform the green recommendation.

Actual prototype captures from presentation slide 4. Color is part of the original experimental conditions.

04 / THE STUDY

Five people.
Three techniques.

A within-subjects VR study asked five adult volunteers to select five targets in a high-density scene with 300 distractors. Each participant used all three techniques.

5adult participants

300distractors in the study scene

4 / 5participants preferred WDIB

Five-participant study · mean ± SD
Technique	Completion time (s) ↓	Attempts per target ↓
Hand collision	37.3 ± 18.9	22.1 ± 9.7
Fixed bubble	15.5 ± 5.4	7.0 ± 1.4
WDIB	19.4 ± 2.3	6.4 ± 0.6

Fixed bubble was fastest on average.

WDIB recorded fewer attempts and lower reported variability on both reported measures.

Participants favored WDIB.

Four of five preferred it. On the reported 1–7 scales, WDIB had mean stability 5.6 and physical demand 1.8.

Original study completion-time chart showing Hand 37.3 seconds, Bubble 15.5 seconds and WDIB 19.4 seconds — Completion time · original chart, slide 10.

Original study attempts chart showing Hand 22.1, Bubble 7.0 and WDIB 6.4 attempts per target — Selection attempts · original chart, slide 11.

Reported measurements ↓Original subjective ratings ↗

UNITY / OPENXR / C#

Explore the prototype.

Build adaptive VR interaction with Unity, OpenXR, and C#.

Code ↗