ID 原文 译文
57858 增加第 4 个上采样特征图，提升高分辨率下密集小物体检测敏感度; Firstly，by adding the fourth upsampling feature map，the detection sensitivity of dense small objects at high resolution is increased.
57859 针对端口固定高宽比特征，利用 k-means 聚类算法重新确定目标候选框个数和高宽比; Secondly，for charac- teristics of the fixed height-width ratio of port，the k-means clustering algorithm is used to re-determine the number of target candidate boxes and the height-width ratio.
57860 提出软非极大值抑制算法，缓解端口靠近且被遮挡情况下引起的漏检、误检; Thirdly，the soft non-maximum suppres- sion algorithm is proposed to alleviate the missed detection and false detection caused by the proximity and occlusion of the port.
57861 针对 4 种疑难生产场景下的端口占用状态完成检测. Finally，four difficult production scenario in the port occupancy state is detec- ted to verify the performance of the improved YOLOv3 algorithm.
57862 实验结果表明，改进后的 YOLOv3 准确率达 90.12% ，相比原 YOLOv3提升了 5.17% . Experiments show that the accuracy of the improved YOLOv3 is 90. 12% ，5. 17% higher than the original YOLOv3.
57863 改进后的算法对于端口类物体具有更高的检测准确率. In conclusion，the im- proved algorithm has higher detection accuracy for port-like objects.
57864 为了充分利用多源异构数据所提供的信息提高推荐准确度，提出一个基于深度学习的混合推荐模型. Considering that Internet information today is diverse and inconsistent in structure，in order to fully utilize the information provided by multi-source heterogeneous data to improve the recommendation accuracy，a hybrid recommendation model based on deep learning was proposed.
57865 该模型融合评分、评论和社交网络数据进行推荐，采用深度学习方法对文本和评分进行特征学习，然后使用社交网络对采样进行约束，从而得到更准确的用户和物品的特征表示. The model makes a rec- ommendation based on combining ratings，review texts and social network data. The model also adopts deep learning to learn features of reviews and ratings，and then uses social network to constraint sam- pling.
57866 实验结果表明，该方法具有较高的准确度. Experiments show that the model is of higher accurate feature representations of users and items.
57867 在公共安全领域查找关键人，需在视频中比对素描进行检索. For public securities，the sketch method is often need to be retrieved in video for seeking key people.