Visual Graphs from Motion (VGfM): Scene understanding with object geometry reasoning
This work addresses scene understanding for computer vision applications, offering an incremental improvement by integrating geometric reasoning from motion into scene graph construction.
The paper tackles the problem of generating scene graphs from single images by leveraging video sequences and multi-view geometry to infer object relationships, achieving results on a newly created dataset for 3D scene graph generation.
Recent approaches on visual scene understanding attempt to build a scene graph -- a computational representation of objects and their pairwise relationships. Such rich semantic representation is very appealing, yet difficult to obtain from a single image, especially when considering complex spatial arrangements in the scene. Differently, an image sequence conveys useful information using the multi-view geometric relations arising from camera motion. Indeed, in such cases, object relationships are naturally related to the 3D scene structure. To this end, this paper proposes a system that first computes the geometrical location of objects in a generic scene and then efficiently constructs scene graphs from video by embedding such geometrical reasoning. Such compelling representation is obtained using a new model where geometric and visual features are merged using an RNN framework. We report results on a dataset we created for the task of 3D scene graph generation in multiple views.