Search papers, labs, and topics across Lattice.
4
0
5
0
Incorporating noisy, in-the-wild imagery into CVL models significantly boosts their performance on clean datasets, challenging traditional reliance on high-end sensors.
Integrating SD-map routes into ego-trajectory prediction yields a remarkable 16.9% reduction in prediction error, proving that high-quality predictions don't always require high-definition maps.
Current online mapping benchmarks are too blunt: a new order-aware metric reveals that detection, not geometric accuracy, is the real bottleneck in state-of-the-art methods.
Achieve geometrically consistent lane-level HD maps from crowd-sourced data using only monocular cameras, consumer-grade GNSS, and IMUs by fusing B-spline representations with a novel Bayesian approach.
