Understanding Look Angle and Target Azimuth Angle To understand the problem, we need to understand two statistics related to remote sensing collects: the look (nadir) angle and the target azimuth angle. This suggests that model is learning to account for look angle in one of these images, but not the other - the predictions from the other image are geospatially shifted by 10–15 pixels, which corresponds to roughly 5–7.5 meters on the ground.
#SPACENET BUSES MANUAL#
Interestingly, the manual labeling was done on the 7 degree imagery (the closest collect to nadir in the dataset), which doesn’t match either of these images perfectly.
Neither set of predictions is perfect, but one clearly matches the manual labels better than the other.Īs you can see, only one of the predicted label sets matches the manually labeled building footprints well. Manually generated building labels (gray) for the image above, overlaid with predictions based on the 29 degree (pink) and 30 degree (green) nadir angle collects. The graph depicting this phenomenon is reproduced below. Strangely, building prediction in images taken at nearly identical look angles - for example, 29 and 30 degrees - produced radically different performance scores. While evaluating the performance of these models, we observed a very unusual phenomenon: “jagged” plot lines for the SpaceNet IoU F1 Score when stratifying evaluation data by look angle.
The SpaceNet Competition Round 4: Off-Nadir Building Detection Challenge has begun! In a recent post, we described competition baseline models that we built, and outlined some of the challenges we encountered while training models to identify building footprints from the dataset. For the first two parts of the series, click here and here.
#SPACENET BUSES SERIES#
This post is part 3 of a series about the SpaceNet 4: Off-Nadir Dataset and Building Detection Challenge. Note: SpaceNet’s mission is to accelerate geospatial machine learning and is supported by the SpaceNet member organizations.
Imagery of the same building in Atlanta, GA at a variety of look angles.