CADENet: Condition-Adaptive Asynchronous Dual-Stream Enhancement Network for Adverse Weather Perception in Autonomous Driving

TL;DR

CADENet is a real-time, training-free system that enhances adverse weather perception in autonomous driving by asynchronously fusing detection and enhancement streams, improving safety-critical detection under challenging conditions.

Contribution

The paper introduces CADENet, a novel asynchronous multi-thread system that enhances weather perception without retraining or extra sensors, addressing evaluation biases in degraded image detection.

Findings

CADENet achieves high recall and F1 scores on snow and rain conditions.

The system sustains approximately 44 FPS regardless of enhancement load.

Evaluation accounts for annotation bias, providing lower bounds on true gains.

Abstract

Adverse weather (rain, fog, sand, and snow) degrades camera-based object detection in autonomous vehicles. Existing enhancement-then-detect approaches stall the safety-critical perception loop, violating hard real-time requirements. Progress on this problem is also constrained by an under-recognized evaluation ceiling: ground truth annotated on degraded images cannot credit a detector that recovers objects the annotators themselves could not see, so a genuinely useful enhancement can register as a near-flat F1 gain. This paper presents CADENet (Condition-Adaptive Asynchronous Dual-stream Enhancement Network), a training-free three-thread system: Thread S (YOLOv11n) delivers detections at full frame rate with zero added latency; Thread Q applies condition-adaptive enhancement (CAPE) and fuses results via entropy-guided NMS (EG-NMS) without blocking Thread S; Thread E provides CLIP zero-shot weather classification, so new weather categories require only a new text prompt, with no labeled data and no retraining. Evaluated on 1327 DAWN images (YOLOv11m, IoU = 0.5, confidence = 0.25), CADENet achieves Recall = 0.0103 (micro), F1 = 0.0230 on snow, and F1 = 0.0038 on rain. We formalize the annotation completeness bias on DAWN-class data, so the reported F1 values are lower bounds on the true gain; recall is the annotation-gap-immune headline metric. Thread S sustains approximately 44 FPS regardless of enhancement load. No model retraining or additional sensor hardware is required.