计算机科学
人工智能
最小边界框
杠杆(统计)
对抗制
目标检测
边距(机器学习)
跳跃式监视
特征(语言学)
模式识别(心理学)
机器学习
领域(数学分析)
背景(考古学)
班级(哲学)
计算机视觉
图像(数学)
数学
古生物学
哲学
数学分析
生物
语言学
作者
Muhammad Akhtar Munir,Muhammad Haris Khan,M. Saquib Sarfraz,Mohsen Ali
标识
DOI:10.1109/tpami.2023.3290135
摘要
Deep learning based object detectors struggle generalizing to a new target domain bearing significant variations in object and background. Most current methods align domains by using image or instance-level adversarial feature alignment. This often suffers due to unwanted background and lacks class-specific alignment. A straightforward approach to promote class-level alignment is to use high confidence predictions on unlabeled domain as pseudo-labels. These predictions are often noisy since model is poorly calibrated under domain shift. In this paper, we propose to leverage model's predictive uncertainty to strike the right balance between adversarial feature alignment and class-level alignment. We develop a technique to quantify predictive uncertainty on class assignments and bounding-box predictions. Model predictions with low uncertainty are used to generate pseudo-labels for self-training, whereas the ones with higher uncertainty are used to generate tiles for adversarial feature alignment. This synergy between tiling around uncertain object regions and generating pseudo-labels from highly certain object regions allows capturing both image and instance-level context during the model adaptation. We report thorough ablation study to reveal the impact of different components in our approach. Results on five diverse and challenging adaptation scenarios show that our approach outperforms existing state-of-the-art methods with noticeable margins.
科研通智能强力驱动
Strongly Powered by AbleSci AI