Search papers, labs, and topics across Lattice.
The authors formulate HyRS-OWOD, an open-world object detection framework that embeds remote sensing data into hyperbolic space to capture the latent hierarchical relationships that Euclidean representations flatten. To tackle unknown-object discovery and catastrophic forgetting, the system decouples semantic-agnostic objectness, leverages hyperbolic embedding radii as uncertainty metrics for unknown detection, and applies hyperbolic metric learning for incremental adaptation. Across three standard remote sensing benchmarks, this geometric shift consistently improves both unknown object recall and incremental learning metrics over Euclidean state-of-the-art baselines.
The radius of a hyperbolic embedding acts as a natural uncertainty signal, allowing open-world detectors to reliably isolate unknown object classes without relying on brittle Euclidean decision boundaries.
Open-world object detection (OWOD) extends closed-set detection by requiring models to identify unknown objects and incrementally learn them once annotations become available. In remote sensing imagery, object categories often exhibit latent hierarchical relationships that may be inadequately represented in the Euclidean spaces commonly adopted by existing methods, limiting unknown-object recall and incremental-learning performance. To address this issue, we investigate hyperbolic geometry for OWOD in remote sensing imagery and propose HyRS-OWOD. To improve unknown object recall, we design a two-step unknown-object discovery mechanism: a Decoupled Objectness Learning (DOL) module that disentangles foreground perception from semantic information to separate foreground proposals from background regions, followed by a Hyperbolic Uncertainty Learning (HUL) component that leverages the radius of hyperbolic embeddings as an uncertainty-aware cue for known-unknown discrimination. For incremental learning, we develop a Hyperbolic Metric Learning (HML) strategy that enhances inter-class separability, facilitating the incorporation of novel categories while mitigating catastrophic forgetting. Experiments on three remote sensing benchmarks demonstrate consistent improvements in unknown recall and incremental learning over state-of-the-art OWOD methods.
