GeoGraph: Learning graph-based multi-view object detection with geometric cues end-to-end

TL;DR

GeoGraph introduces an end-to-end graph neural network approach for multi-view urban object detection and geographic positioning, effectively handling occlusion, viewpoint changes, and multiple views with improved accuracy and efficiency.

Contribution

The paper presents a novel GNN-based method that jointly detects objects, re-identifies instances, and estimates geographic positions from multi-view images in an end-to-end manner.

Findings

Achieves 2-6% higher detection and re-ID precision.

Reduces training time by 8 times.

Demonstrates robustness to occlusion and viewpoint variations.

Abstract

In this paper we propose an end-to-end learnable approach that detects static urban objects from multiple views, re-identifies instances, and finally assigns a geographic position per object. Our method relies on a Graph Neural Network (GNN) to, detect all objects and output their geographic positions given images and approximate camera poses as input. Our GNN simultaneously models relative pose and image evidence, and is further able to deal with an arbitrary number of input views. Our method is robust to occlusion, with similar appearance of neighboring objects, and severe changes in viewpoints by jointly reasoning about visual image appearance and relative pose. Experimental evaluation on two challenging, large-scale datasets and comparison with state-of-the-art methods show significant and systematic improvements both in accuracy and efficiency, with 2-6% gain in detection and re-ID average precision as well as 8x reduction of training time.