Fine-grained image classification focuses on distinguishing subclasses. In food image classification, food's shape, color, and texture are easily influenced by cooking methods, presentation, and lighting. This makes differences among the same kind of food more subtle and harder to capture. To address this, we propose a multimodal framework fusing image and text modalities. The text modality, with distinct statistical properties, data attributes, and relational structures compared to images, enables precise characterization of fine-grained food features, providing complementary semantic information for classification. Specifically, cross-modal knowledge distillation is employed to transfer textual semantics into the image feature space, while contrastive learning is leveraged to discriminate between homogeneous and heterogeneous features across modalities and facilitate feature complementarity. Experiments on benchmark fine-grained food datasets VegFru, FoodX-251, and CNFood show that the proposed method achieves classification accuracies of 97.22 %, 83.25 %, and 84.96 %, respectively, outperforming state-of-the-art (SOTA) methods and validating the effectiveness of the multimodal cross-modal fusion strategy in fine-grained food classification tasks.

应用型本科院校开展职业教育是培养应用型人才的有效途径,是适应产业转型升级的客观要求.通过分析国内外本科职业教育的现状,探索应用型人才供给侧改革的模式与途径,提出应用型本科专业能力与职业能力贯通模型,指导科学制定符合应用型本科职业教育需求的人才培养方案,辅以竞赛驱动、课证融通、产教融合实施方案,培养既有理论功底又有实践技能的复合型专业人才,以适应经济基础的改变和产业技术升级的要求,实现可持续跨越式发展.

本文以高等数学中第一个重要极限为例,将BOPPPS模型的六个环节与超星学习通和课堂相结合,优化教学环节,实现线上与线下混合式教学,激发学生学习兴趣,吸引学生自主地参与课堂活动,并运用所学到的知识解决实际问题.结果表明该模式能促进教学相长,拓展了学生学习的空间和时间,明显提升教学质量,提高了课程的教学效率.

当前,应用型高校计算机专业的学生培养始终存在着课程标准不明确、课程设置失衡、知识和职业技能脱节、实践环节薄弱等问题,教育部在2019年也明确提出课程建设质量需要全面提高,课程是人才培养的核心要素,课程质量直接决定人才培养的质量.全国本科教育必须深化教育改革,必须把教学改革成果落实到课程建设上.

随着互联网技术在教育领域的发展,慕课、爱课程、学习通、智慧课堂等迅速发展并成为不容忽视的教学手段和教育资源.为分析混合式教学方法的教学效果,本文选取同一个学期均开设线性代数的两个专业作为试验班和非试验班,试验班在课程总学时中安排25％的线上学时,由学生通过异步SPOC自主学习,非试验班为全线下授课.教学过程中,两个班级均采用学习通完成课堂签到、随堂知识点练习和课外章节综合测试.通过分析试验班SPOC资源学习情况及两个班级的课外章节综合测试成绩的分布直方图、平均值、及格率、标准差和期末卷面成绩,本文对研究进行总结并提出相关建议,为在线开放式的混合式教学灌入新的思路和依据,实践结果也可以为高校教学改革提供相应的参考.

In most scenarios, a single feature or single point movement is not enough to measure the similarity of rigid body movement. This paper exploited and expanded the method, trajectory similarity measure based on multiple movement features, and proposed a new one to calculate the similarity with multiple features instead of a single feature. Our method integrates the Canny edge detector with Hough transform to measure the similarity of rigid motions. The Canny edge detector is responsible for finding object boundaries in the images. Hough transform detects the line segment and the end points of one rigid body. And then, the similarity of rigid bodies is determined according to the point trajectory.

物联网和人工智能的快速发展给高校嵌入式课程教学带来了很大的挑战,在创新性、技术更新速度和跨学科应用方面都提出了更高的要求.针对嵌入式教学实践中遇到的新问题和新难点,借鉴和融合创客教育模式,对各专业开设的嵌入式课程内容进行重组和融合,构建一个以创新和实践为核心理念的跨学科嵌入式教学体系,在课程设置、教学方法、实验室建设和评价标准上都进行了探索与改革.经教学实践检验能达到预期的效果,大大提升学生创新项目设计和科技竞赛的能力.

Object classification and localization are two significant aspects of object detector based on the Single Shot MultiBox Detector (SSD). In general, the more feature maps there are, the better the object classification performance will be. However, when the information of excessive feature maps are sparse and unnecessary, the performance of object detection is slightly improved or maybe precisely opposite, which is instead harmful to the production of object localization. The performance of object detectors is not only related to the number of feature maps but also relies partly on the bounding box regression and Non-Maximum Suppression (NMS). In this paper, a detector is constructed based on SSD, called Detection with Refined Feature (DRF), involving center map and scale map, the detection loss is reshaped. Our motivation is to improve the accuracy of classification and localization by searching for central points and predicting the scales of the object points. Center map is used to predict the Intersection over Union (IoU) between the prediction box and ground truth box, while scale map considers the relationships among the different scales. Experimental results on both Pascal VOC and MS COCO 2014 instance datasets demonstrate the effectiveness of DRF. Using Darknet53, we achieve an 86.4% mean Average Precision (mAP) on Pascal VOC2007 and an 87.4% mAP on Pascal VOC2007 and VOC2012. On MS COCO, the DRF with ResNet50 still achieves moderate improvement.

With the rapid development of convolutional neural networks, the performance of face detection algorithms has been improved. However, the recall rate of the lightweight face detection algorithm is low when it is applied to the small-scale face and dense face. And high-speed and high-accuracy should be considered at the same time. To solve the problems, we propose a lightweight face detection algorithm based on the YOLOv3 key point. The contributions include: (1) A lightweight face detection algorithm, YOLOFKP, face detector, with scale invariance and enlarged receptive field, is proposed based on YOLOv3. (2) Use mosaic for data preprocessing. (3) The loss function Wing loss and activation function Mish for face detection are used in YOLOv3. (4) The classroom face dataset (CLASSROOMFACE) is made. The experimental results show that YOLOFKP has a fast detection speed and a small number of model parameters. It has robustness for small scale faces and dense faces and effectively improves the recall rate of the lightweight face detection algorithm. The test results on the WIDER FACE dataset show that the face recall rate is 91.19%, and the detection speed of 49 FPS is achieved on a single GTX 1080Ti GPU.

正确识别林区原木运输车辆能有效防止原木被违法运输的异常行为,提高监控管理森林资源的效力.为了解决由于林区道路场景的复杂性,原木端面颜色受光照、湿度等影响使得原木运输车辆识别率较低的问题,将基于YCbCr颜色空间和Hough变换圆检测相结合来识别林区原木运输车辆.同一捆原木端面颜色差异较小,可使用YCbCr颜色特征空间来分割图像,去除背景干扰;图像被转换到RGB空间以去除原木区域二值图像的背景;利用形态学方法统一去除二值化图像的原木缝隙来确定图像边缘;利用Hough变换圆的点线间的对偶性来检测原木运输车辆,降低了噪声的敏感性.实验结果表明,上述方法对成捆裸露在外的原木运输车辆识别率达到了71％以上,鲁棒性和有效性较好.

Background subtraction algorithm is essential for video processing. such as target tracing, gesture recognition and gait recognition. ViBe has been widely used because of easy implementation and high efficiency. However, the algorithm would produce a ghost imaging when the speed of the moving target changes. On the other hand, ViBe is challenged to adapt to the change of environment and by misjudge the shadow as the foreground target. Moreover, it is also inability to handle well the interference caused by camera jitter. Aiming to the deficiencies of ViBe, we propose a new algorithm Gc_IViBe, which takes advantages from both Grabcut and Improved ViBe (IViBe). Based on the ability of IViBe to eliminate ghost imaging, the proposed algorithm utilizes mask in HSV space to remove background shadows. A further improvement of the algorithm in handling cavity problem and camera jitter is achieved by combinating IViBe and Grabcut. The experimental results show that Gc_IViBe performs better than ViBe in Pixel-level measure Precision, Structural measures S-measure and E-measure. This paper also discusses the evaluation methods. The evaluation results of Precision and S-measure in some cases are apparently different from the truth. while E-measure performs relatively better consistent, which capable to accurately evaluate the problems raised in this article.

With the rapid development of convolutional neural networks, the performance of face detection algorithms has been improved. However, the recall rate of the lightweight face detection algorithm is low when it is applied to the small-scale face and dense face. And high-speed and high-accuracy should be considered at the same time. To solve the problems, we propose a lightweight face detection algorithm based on the YOLOv3 key point. The contributions include: (1) A lightweight face detection algorithm, YOLOFKP, face detector, with scale invariance and enlarged receptive field, is proposed based on YOLOv3. (2) Use mosaic for data preprocessing. (3) The loss function Wing loss and activation function Mish for face detection are used in YOLOv3. (4) The classroom face dataset (CLASSROOMFACE) is made. The experimental results show that YOLOFKP has a fast detection speed and a small number of model parameters. It has robustness for small scale faces and dense faces and effectively improves the recall rate of the lightweight face detection algorithm. The test results on the WIDER FACE dataset show that the face recall rate is 91.19%, and the detection speed of 49 FPS is achieved on a single GTX 1080Ti GPU.

在文本分类领域,中文文本需要经过数据处理,将文档表达成计算机可以理解并处理的信息.本文采用TF-IDF作为文本表示方法,针对中文文章的多分类问题,对传统支持向量机进行改进,提出了一种基于特征选择的多类支持向量机分类方法.在中文文章数据集的对比实验结果表明,本文的方法在多分类性能上较优于其他模式识别方法.

This paper introduces a sentiment analysis model based on LSTM (Long Short-Term Memory), which can be used for emotional polarity classification of movie review data.The model consists of five levels, namely input layer, embedding layer, LSTM layer, softmax layer, and output layer.IMDb (Internet Movie Database) dataset is used for training, and SGD (stochastic gradient descent), Adadelta, and RMSprop for optimization.The experiment result shows that Adadelta optimization algorithm can get the best training model and the lowest prediction error rate.

Timber transportation surveillance is the key and important means of forest resource management. In order to surveille timber transportation in the forest road, we carry out timber transportation vehicle detection from the video of bayonet, with which we propose an improvement of SSD (Single Shot Multibox Detector) based on MXNet. We establish a timber transportation vehicle dataset for training and testing the ResNet50 and Darknet53 with three thresholds. As mAP (mean Average Precision) is closely related to IoU (Intersection over Union), the original calculation of which does not consider overlap and disjoint between prediction box and the ground truth, we find it contributes the major part of error-detection with SSD. Inspired by this, we combine a novel metric loss calculation called GIoU (Generalized Intersection over the Union) with SSD to improve mAP, which is proved to be greatly promoted in timber transportation vehicle detection.

针对课程教学学时数有限,而教学内容较多这一矛盾.本文探索一种基于视频课件的引导与督促式教学方法,并以数字信号处理课程教学为例,详细探讨了该教学方法的实施方案.该教学方法的主要思想由"引导"和"督促"两个环节组成.在"引导"环节,要求学生在课前认真观看教师录制的关于知识点的视频课件,确保课前基本了解重要知识点."督促"环节是由课前预习题目来达到目的,要求学生观看完课前知识点之后,完成预习题目.该教学方法能有效地培养学生的自学习惯和提高对新知识点的理解能力.

Traditional text sentiment analysis methods often ignore context information when used in the expression of features. The position of the words in the text makes it difficult to achieve satisfactory results in semantic realization. In recent years, deep learning has obtained good results in text sentiment analysis tasks. Convolutional neural network (CNN) and Recurrent Neural Network(RNN) are two mainstream deep learning algorithms. In this paper, a deep sentiment representation model based on CNNs and long short-term memory recurrent neural network (LSTM) is proposed. The model uses two layers of CNNs to capture the partial features of the text. The model can capture more accurate partial features, after which the features are fed to the LSTM, which can capture the contextual information. Finally, we combine the improved deep learning model with a one-versus-rest training mechanism and apply it to multi-class sentiment classification. We evaluate the proposed model by conducting experiments on datasets. Experimental results demonstrate that the model we designed for multi-class sentiment analysis achieves a accuracy of 78.42% on data set D1 is better than the existing SVMs (support vector machines), CNN, LSTM and CNN-LSTM.

合理的交通控制是缓解城市交通拥堵的重要手段。本文根据车辆的位置、速度、方向等特性构造一个基于元胞自动机的交通控制模型。该模型既可以平均分配路口各个方向的道路上的车流量，也可以分方向动态调整车流量。模型支持基于泊松分布的路口车辆控制，并模拟道路突发事件造成的拥堵对整个区域交通的影响。可视化仿真实验对模型的有效性进行了直观地验证。

在对福建农林大学电子信息工程专业人才培养模式现状进行分析的基础上,提出了电子信息工程专业人才培养模式的一些具体改革建议,旨在培养符合时代发展要求和社会需要的电子信息工程专业人才.

Edge detection based on wavelet transform that can't extract the full and continuous edge of the low-frequency sub-image and just makes use of the high-frequency sub-images may discard some important details of the image and detect poor edges because of image noises effect.Aiming at these problems,this paper proposes a new fusion algorithm based on wavelet transform and mathematical morphology to detect image edges.In the wavelet domain,the low-frequency sub-image edges are detected by mathematical morphology,while the high-frequency sub-images are detected by solving the maximum points of local wavelet coefficient model to restore edges after reducing the noise.Then,the final edge images are obtained through wavelet composition from the fusion edge sub-images according to certain fusion rules.Experimental results show the proposed method can detect image edges not only remove the noise effectively but also enhance the edges and locate edges accurately.