MOA Importance-Driven Selection

Also known as MOAs, Gaze-importance candidate panel

While the user looks around a cluttered scene, gaze builds up an importance score for each object; a hand gesture then opens a 3 × 5 panel of the 15 most-looked-at objects, picked by tilting the palm to a row and tapping a finger for the column.

How it works

The selection half of MOA (multi-object arrangement). Each object's importance grows with how close it is to the gaze point within a 20° foveal cone, discounted when many objects compete, accumulated over time and blended with earlier rounds. Lifting the right hand opens a panel with the 15 highest-scoring objects, normalised in size, in three columns and five rows. Palm pitch selects a row, and tapping the thumb, index or middle finger takes that column's object; tapping several fingers takes several at once. Objects shown but not taken lose their score so new candidates appear next round, and a left thumbs-up ends selection. In a six-session study on a PICO 4 Pro (24 participants, up to 512 objects), the full MOA (selection plus guided placement) was faster with lower workload than a bubble cursor with object proxy and than VVIR.

Papers

Xuehuai Shi, Yuhan Duan, Ziteng Wang, Jian Wu, Zhiwen Shao, Jieming Yin, Lili Wang (2026). MOA: Efficient Scene-Aware Multi-Object Arrangement in VR. IEEE Transactions on Visualization and Computer Graphics 32(2), 2183-2199 doi:10.1109/TVCG.2025.3636062