|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/yolov4-and-tf20/
课程评论：没有评论
**课程名：** YOLO v4 and TF 2.0 **课程概述：** 本课程是计算机视觉领域的第二门课程，将深入探讨两种最先进（SOTA）的目标检测架构：YOLOv4 和 TensorFlow 2.0，并讲解其训练流程。课程还包含一次性数据标注策略，以避免重复标注。 **课程内容：** 课程共分为九个部分： 1. Anaconda 安装 2. 图像数据集重塑 3. 图像数据集标注 4. YOLO 到 PASCAL VOC 格式转换（用于 TF2.0 训练） 5. 在 Google Colab 上进行 YOLOv4 训练和 tflite 转换 6. YOLOv4 Android 部署 7. 在 Google Colab 上进行 SSD Mobilenet TF2.0 训练和 tflite 转换 8. SSD Mobilenet Android 部署 9. YOLOv4 和 SSD 技术细节讲解 **技术细节部分将涵盖：** * **基础概念：** 准确率（Precision）、召回率（Recall）、IoU（Intersection Over Union）、mAP/AP（Mean Average Precision/Average Precision） * **神经网络组件：** Batch Normalization、Residual blocks、Activation function、Max pooling、Feature Pyramid Networks (FPN)、Path Aggregation Network (PAN)、SPP (spatial pyramid pooling layer)、Channel Attention Module (CAM) 和 Spatial Attention Module (SAM) * **YOLOv4 技术细节：** Backbone (CSP), SPP, PAN, SAM, Bag of Freebies (BoF) 和 Bag of Specials (BoS) * **SSD 技术细节：** 架构概述与工作原理、损失函数 * **性能对比：** YOLO vs SSD 的速度和准确率基准测试 **学习目标：** 完成本课程后，学员将能够理解并掌握 YOLOv4 和 TensorFlow 2.0 在目标检测领域的应用，能够进行模型训练、转换和部署，并对相关技术细节有深入了解。
Hi everyone, Welcome to my second course on computer vision. In this course, you will understand the two most latest State Of The Art(SOTA) object detection architecture, which is YOLOv4 and TensorFlow 2.0 and its training pipeline. I also included a one-time labeling strategy, so that you won't have to re-label the image for TensorFlow training. The course is split into 9 parts.Anaconda installation.Image dataset resizing.Image dataset labeling.YOLO to PASCAL VOC conversion for TF2.0 training.YOLOv4 training and tflite conversion on Google Colab.YOLOv4 Android deployment.SSD Mobilenet TF2.0 training and tflite conversion on Google Colab.SSD Mobilenet Android deployment.YOLOv4 and SSD technical details. Which includeBasicsPrecision and RecallIoU(Intersection Over Union)Mean Average Precision/Average Precision(mAP/AP)Batch NormalizationResidual blocksActivation functionMax poolingFeature Pyramid Networks(FPN)Path Aggregation Network (PAN)SPP (spatial pyramid pooling layer)Channel Attention Module(CAM) and Spatial Attention Module (SAM) YOLOv4 - Technical detailsBackboneCross-Stage-Partial-connections (CSP)YOLO with SPPPAN in YOLOv4Spatial Attention Module (SAM) in YOLOv4Bag of freebies (Bof) and Bag of specials (BoS)SSD - Technical detailsArchitecture overview and workingLoss functionsYOLO vs SSDSpeed and accuracy benchmarking
