Google Landmarks Dataset v2 -- A Large-Scale Benchmark for Instance-Level Recognition and Retrieval

3 Apr 2020  ·  Tobias Weyand, Andre Araujo, Bingyi Cao, Jack Sim ·

While image retrieval and instance recognition techniques are progressing rapidly, there is a need for challenging datasets to accurately measure their performance -- while posing novel challenges that are relevant for practical applications. We introduce the Google Landmarks Dataset v2 (GLDv2), a new benchmark for large-scale, fine-grained instance recognition and image retrieval in the domain of human-made and natural landmarks. GLDv2 is the largest such dataset to date by a large margin, including over 5M images and 200k distinct instance labels. Its test set consists of 118k images with ground truth annotations for both the retrieval and recognition tasks. The ground truth construction involved over 800 hours of human annotator work. Our new dataset has several challenging properties inspired by real world applications that previous datasets did not consider: An extremely long-tailed class distribution, a large fraction of out-of-domain test photos and large intra-class variability. The dataset is sourced from Wikimedia Commons, the world's largest crowdsourced collection of landmark photos. We provide baseline results for both recognition and retrieval tasks based on state-of-the-art methods as well as competitive results from a public challenge. We further demonstrate the suitability of the dataset for transfer learning by showing that image embeddings trained on it achieve competitive retrieval performance on independent datasets. The dataset images, ground-truth and metric scoring code are available at https://github.com/cvdfoundation/google-landmark.

PDF Abstract

Datasets


Introduced in the Paper:

Google Landmarks Dataset v2

Used in the Paper:

Places

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Uses Extra
Training Data
Result Benchmark
Landmark Recognition Google Landmarks Dataset v2 (recognition, testing) DELG global+SP microAP 56.35 # 1
Landmark Recognition Google Landmarks Dataset v2 (recognition, validation) DELG global+SP microAP 55.01 # 1
Image Retrieval Google Landmarks Dataset v2 (retrieval, testing) ResNet101+ArcFace GLDv2-train-clean mAP@100 24.15 # 1
Image Retrieval Google Landmarks Dataset v2 (retrieval, validation) ResNet101+ArcFace GLDv2-train-clean mAP@100 22.2 # 1
Image Retrieval ROxford (Hard) ResNet101+ArcFace GLDv2-train-clean mAP 51.6 # 7
Image Retrieval ROxford (Medium) ResNet101+ArcFace GLDv2-train-clean mAP 74.2 # 6
Image Retrieval RParis (Hard) ResNet101+ArcFace GLDv2-train-clean mAP 70.3 # 5
Image Retrieval RParis (Medium) ResNet101+ArcFace GLDv2-train-clean mAP 84.9 # 5

Methods