_强调_ #  **同学帮-文档-代码**  **Classmates-doc-codes** 

![header](./bg.jpg)

#### 介绍 (Introduction)
  本库是 **B站** [同学帮/视觉系](https://space.bilibili.com/202603446) 计算机视觉视觉算法视频配套 `源码及文档`
`同学帮助同学`
This library contains source code and documents of our academic videos, 
which are focus on the fields of `Computer Vision`, `Pattern Recognition` and `Deep Learning`
Please click on [classmates/CV Department](https://space.bilibili.com/202603446) to see our videos.

本站是团队成员（计算机专业、人工智能方向的研究生）内部交流及对外知识共享，由同学们的视角理解算法并动手实验。
Our code is for internal communication among team members and knowledge sharing for those who are intereted in. 
Our purpose is for better understanding of AI algorithm, from the perspective and experiments of our postgraduated students.

部分代码是我们团队成员自己编写，部分来源于其它网站，如果有侵犯到您的权益，请和我们联系，我们将会进行删除。
Some of the code are from us, some of the code comes from open source on the Internet.
If there is any violation of your rights and intellectual property, please contact us to remove it.

*本站代码仅用于学习，用于它用请联系我们*
*Our code is for learning only, please contact us before using it for other purposes*

### 算法列表(Algorithms)
* 我们团队论文讲解, Papers of Our Group
	* [基于形状约束及困难样本挖掘的乳腺超声图像分类, Classification method for samples that are easy to be confused in breast ultrasound images](./OurGroup/HardSample)
	* [乳腺超声双模态数据的协同约束网络, A cooperative suppression network for bimodal data in breast cancer classification](./OurGroup/DualBranch)
	* [Asymmetric Semi-supervised GAN for Breast Ultrasound Image Segmentation](./OurGroup/ASSGAN)
	* [BoundaryFace: A mining framework with noise label self-correction  for Face Recognition](./OurGroup/BoundaryFace)
	* [DSF-Net:Occluded Person Re-identification Based on Dual Structure Features](./OurGroup/DSF_Net)
* [卷积神经网络基础入门实验, Introduction of CNN with Tensorflow](./CNN_Basic)
* [OpenPose: Part Affinity Fields (PAFs)](./PAFs)
* [匈牙利算法, Hungarian algorithm](./Hungary)
* [一种大规模图像检索算法--基于深度局部主注意力特征 Large-Scale Image Retrieval with Attentive Deep Local Features](./DELF)
* [Batch Normlization of CNN](./BatchNormlization)
* [单图像超分辨率重建SRCNN, Single image super resolution reconstruction--SRCNN](./SRCNN)
* [华为云平台ModelArts入门, Introduction of ModelArts](./ModelArts)
* [决策树算法, Decision Tree](./DecisionTree)
* 视频流相关算法, Feature learning for video streams
   * [双流法的融合, Feature fusion methods for two streams](./VideoStream/TwoStreamFusion)
   * [P3D, Pseudo-3D Residual Networks](./VideoStream/P3D)
   * [行为识别, Action recognition with attention](./VideoStream/Action_Recog)
   * [行为识别, Temporal Pyramid Network for Action Recognition](./VideoStream/TPN)
* 目标检测、跟踪, Object detection & tracking
   * [FasterRCNN](./ObjDetection/FasterRCNN)
   * [MaskRCNN](./ObjDetection/MaskRCNN)
   * [用于视频对象检测的内存增强的全局-本地聚合 MEGA for VOD](./ObjDetection/MEGAforVOD)
   * [CornerNet——基于关键点的目标检测, Object detection based on key points](./CornerNet)
   * [Yolo](./ObjDetection)
   * [PG-Net](./ObjDetection/Tracking)
* 行人重识别专辑, Person Re-ID
   * [MGN: Multi-granularities for Re-ID](./Re-ID/MGN)
   * [EANet:Enhancing Alignment for Cross-Domain Re-ID](./Re-ID/EANet)
   * [IDM: An Intermediate Domain Module for Domain Adaptive Person Re-ID](./Re-ID/IDM)
* 人脸识别专辑, Face Recogntion
   * [ArcFace, insight face](./FaceRecogntion)
   * [人脸识别算法发展2022年5月](./FaceRecogntion)
* 非监督学习, UnSupervised Learning
   * [Joint Weakly and Semi-Supervised Breast Ultrasound](./Semi-supervised/JointWeakly)
   * [basic ideas of self-supervised learning](./Semi-supervised/SelfSupervised)
   * [Improving Unsupervised Image Clustering With Robust Learning](./Semi-supervised/UnSupervised)
* 生成对抗网络,Generative Adversarial Networks(./GAN)
* 图像分割，Image Segmentation(./Segmentation)
* Transformer
   * [Introduction of ViT](./Transformer)
   * [Tokens-to-Token ViT](./Transformer)
* 少样本学习 few-shot learning
   * [面向小样本学习的自我训练Self-training for Few-shot Transfer Across Extreme Task Differences](./Few-shot)
* 其它技术、工具等, miscellaneous subjects
   * [Annotation/Label tools for images and videos](./Others/annotationTool)