Cooperative Students: Navigating Unsupervised Domain Adaptation in Nighttime Object Detection

2024-04-02

IEEE International Conference on Multimedia and Expo

AI Summary

To address the significant performance degradation of unsupervised domain adaptation (UDA) for object detection under nighttime low-visibility conditions—thereby undermining autonomous driving robustness—this paper proposes a novel UDA framework. First, a Global-Local Transformation (GLT) module models hierarchical, illumination-invariant features across domains. Second, a Proxy-based Target Consistency (PTC) mechanism enables fine-grained pseudo-label calibration in the unlabeled target domain. Third, an Adaptive IoU-aware Thresholding (AIT) module dynamically optimizes positive sample mining. Built upon Faster R-CNN, our method achieves mAP improvements of 3.0%, 1.9%, and 2.5% on BDD100K, SHIFT, and ACDC benchmarks, respectively, outperforming state-of-the-art approaches. The framework establishes a new paradigm for reliable perception in low signal-to-noise-ratio environments.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Computer Vision: Adversarial Attacks & Robustness
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Unsupervised Domain Adaptation (UDA) has shown significant advancements in object detection under well-lit conditions; however, its performance degrades notably in low-visibility scenarios, especially at night, posing challenges not only for its adaptability in low signal-to-noise ratio (SNR) conditions but also for the reliability and efficiency of automated vehicles. To address this problem, we propose a Cooperative Students (CoS) framework that innovatively employs global-local transformations (GLT) and a proxy-based target consistency (PTC) mechanism to capture the spatial consistency in day- and nighttime scenarios effectively, and thus bridge the significant domain shift across contexts. Building upon this, we further devise an adaptive IoU-informed thresholding (AIT) module to gradually avoid overlooking potential true positives and enrich the latent information in the target domain. Comprehensive experiments show that CoS essentially enhanced UDA performance in low-visibility conditions and surpasses current state-of-the-art techniques, achieving an increase in mAP of 3.0%, 1.9%, and 2.5% on BDD100K, SHIFT, and ACDC datasets, respectively.

Problem

• Improving object detection in low-visibility nighttime conditions
• Bridging domain shift between day and night scenarios
• Enhancing reliability of UDA for automated vehicles

Innovation

• Global-local transformations for spatial consistency
• Proxy-based target consistency mechanism
• Adaptive IoU-informed thresholding module