tang1705 Goto Github PK

ai-cnn

Artificial Intelligence Basic Course Assignment, School of Software Engineering, Beijing Jiaotong University.

ama

Ask me anything!

amusement-park

baidu-rot-validate

利用神经网络（考虑模型、数据集的大小以及模型的感受野）在数据集上进行训练，直接预测需要滑动的距离。由于百度安全验证的角度并不是整数变化的，滑动距离与角度的变化也不是一一对应的，因此相对于预测角度而言，直接预测滑动距离更加准确、便捷。且滑动距离的类别相比较于而言要少，从而使得模型的参数也更少。此外，因为模型的目的是能足够准确地预测滑动距离，从而使得自动化程序模拟验证，因此，应该使得模型尽可能多地在现有数据集上学习。综上，考虑到真实场景的验证图片与获得的图片存在一定的差异（即使是相同的图片，也会受到不同程度的噪声干扰，如水印等），不再对数据集进行训练集与测试集的划分。

baidu-rotnet

bjtu-bachelor-thesis

北京交通大学本科毕业设计（论文）LaTeX 模板（非官方）｜Bachelor Thesis LaTeX Template for Beijing Jiaotong University (unofficial)

bjtu-beamer

北京交通大学 Beamer 主题（非官方）｜Beamer Theme for Beijing Jiaotong University (unofficial)

bjtu-computer-vision

bjtu-deep-learning

bjtu-digital-image-processing

bjtu-graduation-design-2021

Beijing Jiaotong Univ. Undergraduate Graduation Design 2021 | Recipient of the Excellent Undergraduate Graduation Design (Thesis) of Beijing Ordinary Colleges and Universities 🏆

bjtu-information-hiding-and-digital-watermarking

bjtu-report

bjtu-sophomore-mini-semester

基于实时的视频流,利用图像识别功能,自动检测视频中的运动物体。包括后端数据处理和前端 web 页面实时显示结果, 广泛应用在小区入侵监测,智能家居的安防,老人智能看护等领域; 加入人脸识别系统，结合其他的智能产品，实现高度自动化的物联网自动化运行环境。同时支持异常图像上传至云服务器,可以配合手机APP,实现全方位,实时监控的目的。

blog

courses-at-bjtu-scit

curriculum-vitae

data-mining-and-big-data-analysis

This repository is the code of the homework and project for the course named Data Mining and Big Data Analysis in School of Software Engineering, Beijing Jiaotong Unviersity.

duplicate-check-sample

文本查重SDK，可用于论文查重、标书查重、文档查重、作业查重、合同查重、防串标等场景。关联：duplicate check

gf4-based-3d-reconstruction

Part of implementation of paper named Determining Both Surface Position and Orientation in Structured-Light-Based Sensing. Its abstract is Position and orientation profiles are two principal descriptions of shape in space. We describe how a structured light system, coupled with the illumination of a pseudorandom pattern and a suitable choice of feature points, can allow not only the position but also the orientation of individual surface elements to be determined independently. Unlike traditional designs which use the centroids of the illuminated pattern elements as the feature points, the proposed design uses the grid points between the pattern elements instead. The grid points have the essences that their positions in the image data are inert to the effect of perspective distortion, their individual extractions are not directly dependent on one another, and the grid points possess strong symmetry that can be exploited for their precise localization in the image data. Most importantly, the grid lines of the illuminated pattern that form the grid points can aid in determining surface normals. In this paper, we describe how each of the grid points can be labeled with a unique color code, what symmetry they possess and how the symmetry can be exploited for their precise localization at subpixel accuracy in the image data, and how 3D orientation in addition to 3D position can be determined at each of them. Both the position and orientation profiles can be determined with only a single pattern illumination and a single image capture. And the doi of the paper is 10.1109/TPAMI.2009.192.

good-night

happy-reconstruction

轮轨姿态反映着列车运行中轮与轨的接触关系，掌握高铁轮轨姿态是保障高铁安全运营的基础。基于计算机视觉技术的无接触方法可以通过采集运营状态下的轮轨图像，通过三维重建获取轮轨实时姿态。由于高铁运行速度快，轮轨表面光滑，给基于特征点提取的三维重建 (3D Reconstruction) 带来了极大挑战。本项目采用基于空间编码（Space Codification）的编码结构光（coded structured light）的方法，将单幅编码图案（one-shot）投影在轮轨表面，提高特征点的提取和识别精度，并将 De Bruijn 分析与小波变换（wavelet transform）分析相结合，增加了基于特征点的点云提取稠密度，从而实现了单次投影的点云稠密重建。项目完成了从半径95mm的球体提取17W条以上的点云 (Point Cloud) 数据，半径误差0.678mm，实现了对铁轨等多个物体的三维重建，完成了结构光三维重建软件开发，提供了基于主动视觉的三维重建和点云数据可视化展示、编辑的平台。

hello-latex

hello-pytorch

iconbattle

宝石迷阵可以说是消除类游戏的经典之作，游戏的基本规则就是通过在各式各样的宝石中交换相邻的两块宝石，使3个同一颜色的宝石连在一起，就可以消去他们。虽然看起来是很普通而简单的一款游戏，却充分体现了消除类游戏带给玩家的简单中的乐趣。项目在借鉴宝石迷阵游戏基本玩法的前提下进行多方面扩展，既保留了游戏原有的趣味，又体现了开发者的特色，项目名称为 IconBattle。

image-and-video-restoration

image-and-video-understanding

imageduplication

Following work in NLP, the ViT additionally adds a special learned token that is not grounded in the input image. This token serves the purpose to aggregate information into a global image representation which is then used for classification.

intelligent-elderly-care

基于计算机视觉的智慧养老系统通过（模拟）多组摄像头实时拍摄到的画面，用计算机视觉技术实时分析老人的情感、是否有人摔倒、是否有人闯入禁止区域、老人是否有和义工互动、分析是否有陌生人出现并追踪陌生人。一旦上述事件发生，该事件会立即插入到数据库中。这些事件数据被实时地更新在报表中，管理人员因此可以迅速做出反应，从而可以提高管理人员的服务水平和管理能力。本系统分为2部分，分别是 Web 用户界面和基于计算机视觉的摄像头（群组）。本仓库提供了计算机视觉部分的任务及实现。

local-attribution-maps-for-super-resolution

局部归因图(Local Attribution Map， LAM)是一个超分辨率重建任务的可解释性工具，旨在找到低分辨率输入图像中对网络超分结果贡献最强烈的像素。LAM 将跟踪模型使用的信息，并在指定超分结果局部区域的前提下，高亮对超分结果贡献最大的像素。