Large Scale Autonomous Driving Scenarios Clustering with Self-supervised Feature Extraction
This work addresses the need for scalable and unbiased clustering in autonomous driving validation and simulation systems, offering an incremental improvement over existing methods.
The paper tackles the problem of clustering large-scale autonomous driving scenario data by proposing a self-supervised deep learning approach for feature extraction, which avoids biased data representation and surpasses existing methods that rely on handcrafted features, as demonstrated through newly designed evaluation metrics based on data-augmentation.
The clustering of autonomous driving scenario data can substantially benefit the autonomous driving validation and simulation systems by improving the simulation tests' completeness and fidelity. This article proposes a comprehensive data clustering framework for a large set of vehicle driving data. Existing algorithms utilize handcrafted features whose quality relies on the judgments of human experts. Additionally, the related feature compression methods are not scalable for a large data-set. Our approach thoroughly considers the traffic elements, including both in-traffic agent objects and map information. Meanwhile, we proposed a self-supervised deep learning approach for spatial and temporal feature extraction to avoid biased data representation. With the newly designed driving data clustering evaluation metrics based on data-augmentation, the accuracy assessment does not require a human-labeled data-set, which is subject to human bias. Via such unprejudiced evaluation metrics, we have shown our approach surpasses the existing methods that rely on handcrafted feature extractions.