ROVR-Open-Dataset: A Large-Scale Depth Dataset for Autonomous Driving

2025-08-19

AI Summary

Existing deep depth datasets (e.g., KITTI, nuScenes) suffer from limited scene diversity and scale, leading to saturated benchmark performance and hindering progress in foundation models and multimodal depth estimation. To address this, we propose DynamicDrive—the first large-scale monocular depth dataset tailored for dynamic outdoor driving, comprising 20K temporally continuous video frames spanning complex urban environments and long-tail driving conditions. Our method introduces a lightweight vehicle-mounted acquisition pipeline and a sparse yet statistically sufficient LiDAR ground-truth annotation strategy, significantly enhancing scene diversity and depth sparsity while maintaining cost efficiency. Experiments demonstrate substantial performance degradation of state-of-the-art monocular depth models on DynamicDrive—particularly under dynamic occlusions and low-texture regions—validating the dataset’s strong discriminative power and its capacity to expose critical robustness gaps in existing algorithms.

Technology Category

• Computer Vision: 3D Computer Vision
• Application Domains: Mobility, Driving & Flight
• Machine Learning: Deep Neural Architectures and Foundation Models

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Security and Privacy: Large-scale security measurements

Abstract

Depth estimation is a fundamental task for 3D scene understanding in autonomous driving, robotics, and augmented reality. Existing depth datasets, such as KITTI, nuScenes, and DDAD, have advanced the field but suffer from limitations in diversity and scalability. As benchmark performance on these datasets approaches saturation, there is an increasing need for a new generation of large-scale, diverse, and cost-efficient datasets to support the era of foundation models and multi-modal learning. To address these challenges, we introduce a large-scale, diverse, frame-wise continuous dataset for depth estimation in dynamic outdoor driving environments, comprising 20K video frames to evaluate existing methods. Our lightweight acquisition pipeline ensures broad scene coverage at low cost, while sparse yet statistically sufficient ground truth enables robust training. Compared to existing datasets, ours presents greater diversity in driving scenarios and lower depth density, creating new challenges for generalization. Benchmark experiments with standard monocular depth estimation models validate the dataset's utility and highlight substantial performance gaps in challenging conditions, establishing a new platform for advancing depth estimation research.

Problem

• Addressing limitations in diversity and scalability of existing depth datasets
• Providing large-scale diverse cost-efficient data for depth estimation
• Enabling robust training for autonomous driving in dynamic environments

Innovation

• Large-scale diverse depth dataset creation
• Lightweight low-cost acquisition pipeline
• Sparse statistically sufficient ground truth