Mixed Signals V2X Dataset

Submission Dates:
02/18/2025 to 02/14/2026
Citation Author(s):
Katie
Luo
Minh-Quan
Dao
Zhenzhen
Liu
Mark
Campbell
Wei-Lun
Chao
Kilian
Weinberger
Ezio
Malis
Vincent
Fremont
Bharath
Hariharan
Mao
Shan
Stewart
Worrall
Julie Stephany
Berrio Perez
Submitted by:
Katie Luo
Last updated:
Wed, 02/19/2025 - 11:11
DOI:
10.21227/d5bg-cb07
Links:
License:
Creative Commons Attribution

Abstract 

Vehicle-to-everything (V2X) collaborative perception has emerged as a promising solution to address the limitations of single-vehicle perception systems. However, existing V2X datasets are limited in scope, diversity, and quality. To address these gaps, we present Mixed Signals, a comprehensive V2X dataset featuring 45.1k point clouds and 240.6k bounding boxes collected from three connected autonomous vehicles (CAVs) equipped with two different types of LiDAR sensors, plus a roadside unit with dual LiDARs. Our dataset provides precisely aligned point clouds and bounding box annotations across 10 classes, ensuring reliable data for perception training. We provide detailed statistical analysis on the quality of our dataset and extensively benchmark existing V2X methods on it. Mixed Signals V2X Dataset is one of the highest quality, large-scale datasets publicly available for V2X perception research: https://mixedsignalsdataset.cs.cornell.edu/.

Instructions: 

Please visit the Mixed Signals dataset webpage for detailed instructions on how to download and use the dataset. A devkit will be provided for convenience of use.

Comments

,.