Aiming to minimize the total production costs in a single planning period, a nonlinear integer programming model for remanufacturing production plans is established considering the influence of different qualities of ...Aiming to minimize the total production costs in a single planning period, a nonlinear integer programming model for remanufacturing production plans is established considering the influence of different qualities of returns acting on production cost. Three different remanufacturing and discarding strategies are adopted to analyze the change rules of the total production costs. The results returns is greater than indicate that when the number of remanufacturing returns of high the demand, preferentially quality and discarding those of low quality can bring better economic benefits due to manufacturing cost reduction. However, when the number of returns is smaller than the demand, there is no need to consider grading of returns, whereas new demand of remanufacturing. parts are required to satisfy the展开更多
目的观察超像素引导含双流基于局部图像判别器生成对抗网络(SPDD-PatchGAN)用于分割皮肤镜图像中皮肤病变的价值。方法分别于国际皮肤成像合作组织(ISIC)2016数据集及Human Against Machine with 10000 training images(HAM10000)数据...目的观察超像素引导含双流基于局部图像判别器生成对抗网络(SPDD-PatchGAN)用于分割皮肤镜图像中皮肤病变的价值。方法分别于国际皮肤成像合作组织(ISIC)2016数据集及Human Against Machine with 10000 training images(HAM10000)数据集中收集1279及10015幅皮肤病变患者的皮肤镜图像,以数据集中的手动分割结果为参考标准,利用包含多尺度上下文提取模块(MCEM)的残差注意力UNet(RA-UNet)为生成器,采用基于局部图像的超像素引导的双流判别策略为判别器构建SPDD-PatchGAN并以之分割图像中的皮肤病变;与深度卷积生成对抗网络(DCGAN)、UNet、Attention-UNet、上下文编码器网络(CENet)、上下文金字塔融合网络(CPFNet)和生成对抗模型双流生成对抗网络(DAGAN)对比,采用平均交并比(mIoU)、准确率(Accuracy)、召回率(Recall)评价SPDD-PatchGAN的分割效能。结果SPDD-PatchGAN分割皮肤病变的整体效果较佳,且其mIoU、Accuracy及Recall均优于DCGAN、UNet、Attention-UNet、CENet、CPFNet及DAGAN。结论SPDD-PatchGAN可有效分割皮肤镜图像中的皮肤病变。展开更多
基金The National Natural Science Foundation of China(No.70671022)
文摘Aiming to minimize the total production costs in a single planning period, a nonlinear integer programming model for remanufacturing production plans is established considering the influence of different qualities of returns acting on production cost. Three different remanufacturing and discarding strategies are adopted to analyze the change rules of the total production costs. The results returns is greater than indicate that when the number of remanufacturing returns of high the demand, preferentially quality and discarding those of low quality can bring better economic benefits due to manufacturing cost reduction. However, when the number of returns is smaller than the demand, there is no need to consider grading of returns, whereas new demand of remanufacturing. parts are required to satisfy the
文摘目的观察超像素引导含双流基于局部图像判别器生成对抗网络(SPDD-PatchGAN)用于分割皮肤镜图像中皮肤病变的价值。方法分别于国际皮肤成像合作组织(ISIC)2016数据集及Human Against Machine with 10000 training images(HAM10000)数据集中收集1279及10015幅皮肤病变患者的皮肤镜图像,以数据集中的手动分割结果为参考标准,利用包含多尺度上下文提取模块(MCEM)的残差注意力UNet(RA-UNet)为生成器,采用基于局部图像的超像素引导的双流判别策略为判别器构建SPDD-PatchGAN并以之分割图像中的皮肤病变;与深度卷积生成对抗网络(DCGAN)、UNet、Attention-UNet、上下文编码器网络(CENet)、上下文金字塔融合网络(CPFNet)和生成对抗模型双流生成对抗网络(DAGAN)对比,采用平均交并比(mIoU)、准确率(Accuracy)、召回率(Recall)评价SPDD-PatchGAN的分割效能。结果SPDD-PatchGAN分割皮肤病变的整体效果较佳,且其mIoU、Accuracy及Recall均优于DCGAN、UNet、Attention-UNet、CENet、CPFNet及DAGAN。结论SPDD-PatchGAN可有效分割皮肤镜图像中的皮肤病变。
