PACE: Polar Axis-Conditioned Estimation for PairUAV Relative Localization
Computer Vision and Pattern Recognition
Summary
The authors looked at how two drones can figure out their positions relative to each other using images. They found that trying to predict both direction (heading) and distance (range) with the same method doesn’t work well because each needs different clues from the data. So, they created a new way that treats direction and distance separately while still sharing some information, improving accuracy. Their approach was tested on a hidden dataset and showed better results than previous methods. They also made their code and tools publicly available for others to use.
Authors
Ze Rong
Abstract
PairUAV relative localization maps two UAV images to a polar navigation command. Although heading and range share the same pairwise pose context, treating them as homogeneous coordinates forces both outputs to use the same decoder evidence and optimization state. Controlled readout probes reveal a different structure: the two axes favor different decoder-depth combinations, their best checkpoints disagree on 80.8% of a validation trajectory, and range errors exhibit a distinct high-error tail. We introduce method, Polar Axis-Conditioned Estimation, which retains a shared Reloc3r-style pair representation while assigning axis-specific readout interfaces. Heading uses mid/late relational evidence, whereas range remains attached to a direct late metric path. On the official hidden test, the strongest released raw predictor scores 0.002460; the complementary PAAER predictor scores 0.002514 with a slightly lower angle error. Deterministic challenge packaging, reported separately from learned estimation, yields the final score of 0.001874. Code, checkpoints, predictions, and reconstruction tools are available at https://github.com/zerong7777-boop/PairUAV-PACE.