Snap Angle Prediction for 360$^{\circ}$ Panoramas
This addresses the challenge of content-aware projection for 360° panoramas, which is incremental as it builds on existing projection models with a new method.
The paper tackles the problem of visualizing 360° panoramas in 2D by predicting optimal rotation angles to reduce distortions, resulting in more visually pleasing panoramas with 5x less computation than the baseline.
360$^{\circ}$ panoramas are a rich medium, yet notoriously difficult to visualize in the 2D image plane. We explore how intelligent rotations of a spherical image may enable content-aware projection with fewer perceptible distortions. Whereas existing approaches assume the viewpoint is fixed, intuitively some viewing angles within the sphere preserve high-level objects better than others. To discover the relationship between these optimal snap angles and the spherical panorama's content, we develop a reinforcement learning approach for the cubemap projection model. Implemented as a deep recurrent neural network, our method selects a sequence of rotation actions and receives reward for avoiding cube boundaries that overlap with important foreground objects. We show our approach creates more visually pleasing panoramas while using 5x less computation than the baseline.