An uncertain multi-objective programming problem is a special type of mathematical multi-objective programming involving uncertain variables. This type of problem is important because there are several uncertain varia...An uncertain multi-objective programming problem is a special type of mathematical multi-objective programming involving uncertain variables. This type of problem is important because there are several uncertain variables in real-world problems.Therefore, research on the uncertain multi-objective programming problem is highly relevant, particularly those problems whose objective functions are correlated. In this paper, an approach that solves an uncertain multi-objective programming problem under the expected-variance value criterion is proposed. First, we define the basic framework of the approach and review concepts such as a Pareto efficient solution and expected-variance value criterion using an order relation between various uncertain variables.Second, the uncertain multi-objective problem is converted into an uncertain single-objective programming problem via a linear weighted method or ideal point method. Then the problem is transformed into a deterministic single objective programming problem under the expected-variance value criterion. Third, four lemmas and two theorems are proved to illustrate that the optimal solution of the deterministic single-objective programming problem is an efficient solution to the original uncertainty problem. Finally, two numerical examples are presented to validate the effectiveness of the proposed approach.展开更多
构建大规模茶芽目标检测数据集是一项耗时且繁琐的任务,为了降低数据集构建成本,探索少量标注样本的算法尤为必要。本文提出了YSVD-Tea(YOLO singular value decomposition for tea bud detection)算法,通过将预训练模型中的基础卷积替...构建大规模茶芽目标检测数据集是一项耗时且繁琐的任务,为了降低数据集构建成本,探索少量标注样本的算法尤为必要。本文提出了YSVD-Tea(YOLO singular value decomposition for tea bud detection)算法,通过将预训练模型中的基础卷积替换为3个连续的矩阵结构,实现了对YOLOX算法结构的重构。通过维度变化和奇异值分解操作,将预训练权重转换为与重构算法结构相对应的权重,从而将需要进行迁移学习的权重和需要保留的权重分离开,实现保留预训练模型先验信息的目的。在3种不同数量的数据集上分别进行了训练和验证。在最小数量的1/3数据集上,YSVD-Tea算法相较于改进前的YOLOX算法,mAP提高20.3个百分点。对比测试集与训练集的性能指标,YSVD-Tea算法在测试集与训练集的mAP差距仅为21.9%,明显小于YOLOX的40.6%和Faster R-CNN的55.4%。在数量最大的数据集上,YOLOX算法精确率、召回率、F1值、mAP分别为86.4%、87.0%、86.7%和88.3%,相较于对比算法均最高。YSVD-Tea在保证良好性能的同时,能够更好地适应少量标注样本的茶芽目标检测任务。展开更多
近年来,无人机因体积小、灵活性好等优势被广泛应用在车辆跟踪领域。当无人机在高空飞行时,其捕捉的图像中车辆目标存在像素点少、拥挤以及被遮挡的情况。现有的多目标跟踪研究方法在车辆被遮挡过程中发生非线性运动时,使用卡尔曼滤波预...近年来,无人机因体积小、灵活性好等优势被广泛应用在车辆跟踪领域。当无人机在高空飞行时,其捕捉的图像中车辆目标存在像素点少、拥挤以及被遮挡的情况。现有的多目标跟踪研究方法在车辆被遮挡过程中发生非线性运动时,使用卡尔曼滤波预测,会出现车辆位置预测不准确的问题。为了解决这些问题,采用先检测后跟踪(tracking by detection,TBD)范式,对YOLOv8检测算法进行改进,在网络结构中引入了BiFormer稀疏动态注意力模块,用于提取小目标特征信息。同时使用轻量级上采样算子CARAFE替换原最近邻插值上采样,减少上采样过程中小目标特征丢失的问题。提出一种轻量化跟踪模型FA-SORT,针对SORT算法提出三点改进:改进KF、添加速度方向一致性匹配和检测值匹配。在自制地组合了多个车辆数据集上验证改进的YOLOv8算法。实验结果表明,与YOLOv8相比,精确率(precision)提高了0.97%,召回率(recall)提高了0.898%。对所提出的FA-SORT算法使用UAVDT数据集进行验证,结果表明,与现有的多目标跟踪算法相比,HOTA指标首个达到70.05%,IDF1达到87.45%,跟踪速度达到29.93 FPS。验证了FA-SORT跟踪算法在多车辆跟踪任务中的优越性。展开更多
基金supported by the National Natural Science Foundation of China(71601183 71571190)
文摘An uncertain multi-objective programming problem is a special type of mathematical multi-objective programming involving uncertain variables. This type of problem is important because there are several uncertain variables in real-world problems.Therefore, research on the uncertain multi-objective programming problem is highly relevant, particularly those problems whose objective functions are correlated. In this paper, an approach that solves an uncertain multi-objective programming problem under the expected-variance value criterion is proposed. First, we define the basic framework of the approach and review concepts such as a Pareto efficient solution and expected-variance value criterion using an order relation between various uncertain variables.Second, the uncertain multi-objective problem is converted into an uncertain single-objective programming problem via a linear weighted method or ideal point method. Then the problem is transformed into a deterministic single objective programming problem under the expected-variance value criterion. Third, four lemmas and two theorems are proved to illustrate that the optimal solution of the deterministic single-objective programming problem is an efficient solution to the original uncertainty problem. Finally, two numerical examples are presented to validate the effectiveness of the proposed approach.
文摘构建大规模茶芽目标检测数据集是一项耗时且繁琐的任务,为了降低数据集构建成本,探索少量标注样本的算法尤为必要。本文提出了YSVD-Tea(YOLO singular value decomposition for tea bud detection)算法,通过将预训练模型中的基础卷积替换为3个连续的矩阵结构,实现了对YOLOX算法结构的重构。通过维度变化和奇异值分解操作,将预训练权重转换为与重构算法结构相对应的权重,从而将需要进行迁移学习的权重和需要保留的权重分离开,实现保留预训练模型先验信息的目的。在3种不同数量的数据集上分别进行了训练和验证。在最小数量的1/3数据集上,YSVD-Tea算法相较于改进前的YOLOX算法,mAP提高20.3个百分点。对比测试集与训练集的性能指标,YSVD-Tea算法在测试集与训练集的mAP差距仅为21.9%,明显小于YOLOX的40.6%和Faster R-CNN的55.4%。在数量最大的数据集上,YOLOX算法精确率、召回率、F1值、mAP分别为86.4%、87.0%、86.7%和88.3%,相较于对比算法均最高。YSVD-Tea在保证良好性能的同时,能够更好地适应少量标注样本的茶芽目标检测任务。
文摘近年来,无人机因体积小、灵活性好等优势被广泛应用在车辆跟踪领域。当无人机在高空飞行时,其捕捉的图像中车辆目标存在像素点少、拥挤以及被遮挡的情况。现有的多目标跟踪研究方法在车辆被遮挡过程中发生非线性运动时,使用卡尔曼滤波预测,会出现车辆位置预测不准确的问题。为了解决这些问题,采用先检测后跟踪(tracking by detection,TBD)范式,对YOLOv8检测算法进行改进,在网络结构中引入了BiFormer稀疏动态注意力模块,用于提取小目标特征信息。同时使用轻量级上采样算子CARAFE替换原最近邻插值上采样,减少上采样过程中小目标特征丢失的问题。提出一种轻量化跟踪模型FA-SORT,针对SORT算法提出三点改进:改进KF、添加速度方向一致性匹配和检测值匹配。在自制地组合了多个车辆数据集上验证改进的YOLOv8算法。实验结果表明,与YOLOv8相比,精确率(precision)提高了0.97%,召回率(recall)提高了0.898%。对所提出的FA-SORT算法使用UAVDT数据集进行验证,结果表明,与现有的多目标跟踪算法相比,HOTA指标首个达到70.05%,IDF1达到87.45%,跟踪速度达到29.93 FPS。验证了FA-SORT跟踪算法在多车辆跟踪任务中的优越性。
