Technology Category
Intelligent Robots: Multimodal Perception & Sensor FusionComputer Vision: Low Level & Physics-based VisionKnowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
ReaLiTy and LADS: A Unified Framework and Dataset Suite for LiDAR Adaptation Across Sensors and Adverse Weather Conditions
🤖 AI Summary
Existing LiDAR perception methods suffer from insufficient robustness under varying sensor configurations and adverse weather conditions, compounded by a lack of physically consistent cross-domain data. To address this, this work proposes ReaLiTy, a unified physics-informed framework that integrates physics-guided geometric and radiometric degradation models, a learning-based intensity generation module, and a one-to-one transformation mechanism grounded in real-world measurements. ReaLiTy enables, for the first time, the generation of physically consistent LiDAR point clouds across different sensors and weather conditions. The accompanying LADS dataset suite provides rigorously aligned multi-domain observations, significantly enhancing cross-domain consistency and the realism of weather-induced effects, thereby establishing a reproducible benchmark for simulation-driven LiDAR perception research.
Technology Category
Application Category
📝 Abstract
Reliable LiDAR perception requires robustness across sensors, environments, and adverse weather. However, existing datasets rarely provide physically consistent observations of the same scene under varying sensor configurations and weather conditions, limiting systematic analysis of domain shifts. This work presents ReaLiTy, a unified physics-informed framework that transforms LiDAR data to match target sensor specifications and weather conditions. The framework integrates physically grounded cues with a learning-based module to generate realistic intensity patterns, while a physics-based weather model introduces consistent geometric and radiometric degradations. Building on this framework, we introduce the LiDAR Adaptation Dataset Suite (LADS), a collection of physically consistent, transformation-ready point clouds with one-to-one correspondence to original datasets. Experiments demonstrate improved cross-domain consistency and realistic weather effects. ReaLiTy and LADS provide a reproducible foundation for studying LiDAR adaptation and simulation-driven perception in intelligent transportation systems.
Problem
Research questions and friction points this paper is trying to address.
LiDAR adaptation
domain shift
adverse weather
sensor variation
physical consistency
Innovation
Methods, ideas, or system contributions that make the work stand out.
LiDAR adaptation
physics-informed framework
adverse weather simulation
cross-sensor consistency
realistic point cloud generation
🔎 Similar Papers
No similar papers found.
💼 Related Jobs
No related jobs found.
