Enhanced Drift-Aware Computer Vision Architecture for Autonomous Driving

2025-08-25

AI Summary

To address significant performance degradation in autonomous driving object detection caused by data drift under adverse conditions—such as poor weather and low illumination—this paper proposes a drift-aware dual-modal vision architecture. The method innovatively couples a YOLOv8 detector with a five-layer CNN-based verification module, establishing a cascaded “fast detection–fine-grained verification” mechanism. Joint training is performed on large-scale synthetic road-scene data to enhance robustness against unknown distribution shifts. Experimental results demonstrate over 90% improvement in detection accuracy across multiple drift-augmented test sets, substantially mitigating performance degradation and enhancing system safety and stability. This work contributes a scalable architectural paradigm and a practical training strategy for robust perception in open-world autonomous driving environments.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Adversarial Learning & Robustness
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

The use of computer vision in automotive is a trending research in which safety and security are a primary concern. In particular, for autonomous driving, preventing road accidents requires highly accurate object detection under diverse conditions. To address this issue, recently the International Organization for Standardization (ISO) released the 8800 norm, providing structured frameworks for managing associated AI relevant risks. However, challenging scenarios such as adverse weather or low lighting often introduce data drift, leading to degraded model performance and potential safety violations. In this work, we present a novel hybrid computer vision architecture trained with thousands of synthetic image data from the road environment to improve robustness in unseen drifted environments. Our dual mode framework utilized YOLO version 8 for swift detection and incorporated a five-layer CNN for verification. The system functioned in sequence and improved the detection accuracy by more than 90% when tested with drift-augmented road images. The focus was to demonstrate how such a hybrid model can provide better road safety when working together in a hybrid structure.

Problem

• Addressing data drift in autonomous driving object detection
• Improving computer vision robustness under adverse conditions
• Enhancing safety through hybrid architecture for drift scenarios

Innovation

• Hybrid computer vision architecture with synthetic data training
• Dual mode framework using YOLOv8 and five-layer CNN
• Sequential processing for enhanced drift robustness