ERUP-YOLO: Enhancing Object Detection Robustness for Adverse Weather Condition by Unified Image-Adaptive Processing

2024-11-05

arXiv.org

AI Summary

To address the significant degradation in object detection performance under adverse weather conditions—such as fog and low illumination—this paper proposes an image-adaptive detection framework. The core innovation lies in a dual differentiable filter design: a Bézier-curve pixel-wise (BPW) filter for nonlinear pixel-level enhancement, and a kernel-based local (KBL) filter for spatially adaptive convolution; both are seamlessly integrated into the YOLOv3 pipeline and jointly optimized in an end-to-end trainable manner. Furthermore, we introduce a domain-agnostic BPW data augmentation strategy that requires no manual parameter tuning or scene-specific customization. Extensive experiments on multiple adverse-weather benchmarks demonstrate consistent superiority over state-of-the-art methods, achieving mAP gains of 3.2–5.8%. The framework exhibits strong generalization, robustness to diverse weather degradations, and real-time inference speed (≥23 FPS).

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Adversarial Learning & Robustness
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

We propose an image-adaptive object detection method for adverse weather conditions such as fog and low-light. Our framework employs differentiable preprocessing filters to perform image enhancement suitable for later-stage object detections. Our framework introduces two differentiable filters: a B'ezier curve-based pixel-wise (BPW) filter and a kernel-based local (KBL) filter. These filters unify the functions of classical image processing filters and improve performance of object detection. We also propose a domain-agnostic data augmentation strategy using the BPW filter. Our method does not require data-specific customization of the filter combinations, parameter ranges, and data augmentation. We evaluate our proposed approach, called Enhanced Robustness by Unified Image Processing (ERUP)-YOLO, by applying it to the YOLOv3 detector. Experiments on adverse weather datasets demonstrate that our proposed filters match or exceed the expressiveness of conventional methods and our ERUP-YOLO achieved superior performance in a wide range of adverse weather conditions, including fog and low-light conditions.

Problem

• Adverse Weather Conditions
• Image Recognition
• Object Detection

Innovation

• Adaptive Image Processing
• Bezier Curves Integration
• Flexible Data Augmentation