CVM-Net: Cross-View Matching Network for Image-Based Ground-to-Aerial Geo-Localization

By Sixing Hu, Mengdan Feng, Rang M. H. Nguyen, Gim Hee Lee. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018

395 citations recorded by the source corpus.

The problem of localization on a geo-referenced aerial/satellite map given a query ground view image remains challenging due to the drastic change in viewpoint that causes traditional image descriptors based matching to fail. We leverage on the recent success of deep learning to propose the CVM-Net for the cross-view image-based ground-to-aerial geo-localization task. Specifically, our network is based on the Siamese architecture to do metric learning for the matching task. We first use the fully convolutional layers to extract local image features, which are then encoded into global image descriptors using the powerful NetVLAD. As part of the training procedure, we also introduce a simple yet effective weighted soft margin ranking loss function that not only speeds up the training convergence but also improves the final matching accuracy. Experimental results show that our proposed network significantly outperforms the state-of-the-art approaches on two existing benchmarking datasets. Our code and models are publicly available on the project website1.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/f7585d0c4eb361c49de21a84d3e7c66b12991a1f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

Each Part Matters: Local Patterns Facilitate Cross-View Geo-Localization (2020, methodology): “CVM-Net [7] employs a weighted soft margin ranking loss, which not only speeds up the training convergence but also improves the retrieval accuracy.”
University-1652: A Multi-view Multi-source Benchmark for Drone-based Geo-localization (2020, methodology): “e the high-level feature space but pay attention to different low-level patterns of inputs from Methods R@1 R@5 R@10 R@Top1% Workman [31] - - - 34.40 Zhai [34] - - - 43.20 Vo [29] - - - 63.70 CVM-Net [11] 18.80 44.42 57.47 91.54 Orientation [16]† 27.15 54.66 67.54 93.91 Ours 43.91 66.38 74.58 91.78 ”
VIGOR: Cross-View Image Geo-localization beyond One-to-one Retrieval (2020, methodology): “For the baseline, we adopt the widely used triplet-like loss proposed in [7]:”
Lending Orientation to Neural Networks for Cross-View Geo-Localization (2019, methodology): “[11] proposed to use NetVLAD [5] as feature aggregation on top of the VGG [28] architecture pre-trained on ImageNet dataset [7], and obtained the state-of-the-art performance.”
Spatial-Aware Feature Aggregation for Image based Cross-View Geo-Localization (2019, methodology): “Similar to [5], we employ a weighted soft-margin triplet loss as our objective: L = log(1 + eγ(dpos−dneg)), (3)”