Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

SCOPE: A Synthetic Multi-Modal Dataset for Collective Perception Including Physical-Correct Weather Conditions

Jörg Gamerdinger, Sven Teufel, Patrick Schulz, Stephan Amann, Jan-Patrick Kirchner, Oliver Bringmann

TL;DR

The paper introduces SCOPE, a synthetic, multimodal dataset for collective perception that integrates physically accurate weather effects for both camera and LiDAR sensors, plus realistic LiDAR models. It offers 44 diverse driving scenarios across two digital-twin maps (Karlsruhe and Tübingen) with up to 24 collaborative agents and a full sensor suite (RGB, SemSeg, BEV, and three LiDARs). Ground truth includes detailed 3D pose, velocity, and segmentation labels, enabling evaluation of 2D/3D object detection and semantic segmentation under adverse conditions. The dataset is complemented by a Python toolkit and a comprehensive benchmark, supporting robust training and cross-domain evaluation for V2X-based collective perception and domain adaptation studies.

Abstract

Collective perception has received considerable attention as a promising approach to overcome occlusions and limited sensing ranges of vehicle-local perception in autonomous driving. In order to develop and test novel collective perception technologies, appropriate datasets are required. These datasets must include not only different environmental conditions, as they strongly influence the perception capabilities, but also a wide range of scenarios with different road users as well as realistic sensor models. Therefore, we propose the Synthetic COllective PErception (SCOPE) dataset. SCOPE is the first synthetic multi-modal dataset that incorporates realistic camera and LiDAR models as well as parameterized and physically accurate weather simulations for both sensor types. The dataset contains 17,600 frames from over 40 diverse scenarios with up to 24 collaborative agents, infrastructure sensors, and passive traffic, including cyclists and pedestrians. In addition, recordings from two novel digital-twin maps from Karlsruhe and Tübingen are included. The dataset is available at https://ekut-es.github.io/scope

SCOPE: A Synthetic Multi-Modal Dataset for Collective Perception Including Physical-Correct Weather Conditions

TL;DR

The paper introduces SCOPE, a synthetic, multimodal dataset for collective perception that integrates physically accurate weather effects for both camera and LiDAR sensors, plus realistic LiDAR models. It offers 44 diverse driving scenarios across two digital-twin maps (Karlsruhe and Tübingen) with up to 24 collaborative agents and a full sensor suite (RGB, SemSeg, BEV, and three LiDARs). Ground truth includes detailed 3D pose, velocity, and segmentation labels, enabling evaluation of 2D/3D object detection and semantic segmentation under adverse conditions. The dataset is complemented by a Python toolkit and a comprehensive benchmark, supporting robust training and cross-domain evaluation for V2X-based collective perception and domain adaptation studies.

Abstract

Collective perception has received considerable attention as a promising approach to overcome occlusions and limited sensing ranges of vehicle-local perception in autonomous driving. In order to develop and test novel collective perception technologies, appropriate datasets are required. These datasets must include not only different environmental conditions, as they strongly influence the perception capabilities, but also a wide range of scenarios with different road users as well as realistic sensor models. Therefore, we propose the Synthetic COllective PErception (SCOPE) dataset. SCOPE is the first synthetic multi-modal dataset that incorporates realistic camera and LiDAR models as well as parameterized and physically accurate weather simulations for both sensor types. The dataset contains 17,600 frames from over 40 diverse scenarios with up to 24 collaborative agents, infrastructure sensors, and passive traffic, including cyclists and pedestrians. In addition, recordings from two novel digital-twin maps from Karlsruhe and Tübingen are included. The dataset is available at https://ekut-es.github.io/scope
Paper Structure (13 sections, 1 equation, 4 figures, 2 tables)

This paper contains 13 sections, 1 equation, 4 figures, 2 tables.

Table of Contents

  1. INTRODUCTION
  2. RELATED WORK
  3. SCOPE DATASET
  4. Simulation Setup
  5. Scenarios
  6. Sensor Setup
  7. Environmental Variation
  8. Image Augmentation
  9. Point Cloud Augmentation
  10. Data Structure
  11. Toolkit
  12. BENCHMARK
  13. CONCLUSION & OUTLOOK

Figures (4)

  • Figure 1: Exemplary scenes from Tübingen (top) and Karlsruhe (bottom)
  • Figure 2: Sensor Setup of the CAVs
  • Figure 3: Exemplary scenes including the weather simulation for camera data (a) and LiDAR data (b) with clear weather (left), rain (mid) and fog (right)
  • Figure 4: Statistics of the SCOPE dataset