Learning-based Relational Object Matching Across Views
This addresses the challenge of object-level scene understanding for intelligent robots and perception tasks like scene reconstruction, though it is incremental as it builds on existing object-level matching approaches.
The paper tackles the problem of matching objects across large viewpoint changes in RGB images by proposing a learning-based approach that combines local keypoints with novel object-level features, achieving improved performance over pure keypoint-based methods for such changes.
Intelligent robots require object-level scene understanding to reason about possible tasks and interactions with the environment. Moreover, many perception tasks such as scene reconstruction, image retrieval, or place recognition can benefit from reasoning on the level of objects. While keypoint-based matching can yield strong results for finding correspondences for images with small to medium view point changes, for large view point changes, matching semantically on the object-level becomes advantageous. In this paper, we propose a learning-based approach which combines local keypoints with novel object-level features for matching object detections between RGB images. We train our object-level matching features based on appearance and inter-frame and cross-frame spatial relations between objects in an associative graph neural network. We demonstrate our approach in a large variety of views on realistically rendered synthetic images. Our approach compares favorably to previous state-of-the-art object-level matching approaches and achieves improved performance over a pure keypoint-based approach for large view-point changes.