Google MediaPipe 演示Demo（脸部、手指、肢体动作识别，及即时运动跟踪）

完整内容请访问：https://blog.csdn.net/chitiantong/article/details/114937760

 　　 

文章目录
本篇文章充当搬运工，赚点下载积分。
1.AutoFlip Saliency-aware Video Cropping
2.Box Tracking
3.Dataset Preparation with MediaSequence
4.Face Detection 人脸检测
5.Face Mesh 人脸网格
6.Hair Segmentation 头发分割
7.Hands 手部检测和跟踪
8.Holistic 全面跟踪
9.Instant Motion Tracking 即时运动跟踪
10.Iris 虹膜
11.KNIFT (Template-based Feature Matching)
12.Object Detection
13.Objectron (3D Object Detection) 三维目标检测
14.Pose 姿势跟踪
本篇文章充当搬运工，赚点下载积分。
1.AutoFlip Saliency-aware Video Cropping
自动翻转显著性感知视频裁剪

AutoFlip is an automatic video cropping pipeline built on top of MediaPipe. This example focuses on demonstrating how to use AutoFlip to convert an input video to arbitrary aspect ratios.

AutoFlip是构建在MediaPipe之上的自动视频裁剪管道。本示例重点演示如何使用AutoFlip将输入视频转换为任意纵横比。

2.Box Tracking
下载 objecttrackinggpu.apk

3.Dataset Preparation with MediaSequence
用MediaSequence准备数据集

MediaPipe is a useful and general framework for media processing that can assist with research, development, and deployment of ML models. This example focuses on development by demonstrating how to prepare video data for training a TensorFlow model.

The MediaSequence library provides an extensive set of tools for storing data in TensorFlow.SequenceExamples. SequenceExamples provide matched semantics to most video tasks and are efficient to use with TensorFlow. The sequence semantics allow for a variable number of annotations per frame, which is necessary for tasks like video object detection, but very difficult to encode in TensorFlow.Examples. The goal of MediaSequence is to simplify working with SequenceExamples and to automate common preparation tasks. Much more information is available about the MediaSequence pipeline, including how to use it to process new data sets, in the documentation of MediaSequence.

MediaPipe是一个有用的、通用的媒体处理框架，可以帮助研究、开发和部署ML模型。本例通过演示如何准备用于训练张量流模型的视频数据来关注开发。

MediaSequence库提供了一套广泛的工具，用于将数据存储在TensorFlow.Sequence示例. SequenceExamples为大多数视频任务提供匹配的语义，并且可以有效地与TensorFlow一起使用。序列语义允许每帧具有可变数量的注释，这对于视频对象检测之类的任务是必需的，但是很难在其中进行编码TensorFlow.示例. MediaSequence的目标是简化使用SequenceExamples的工作，并自动化常见的准备任务。MediaSequence文档中提供了有关MediaSequence管道的更多信息，包括如何使用它处理新的数据集。

4.Face Detection 人脸检测
下载 facedetectiongpu.apk

5.Face Mesh 人脸网格
下载 faceeffect.apk

 

6.Hair Segmentation 头发分割
下载 hairsegmentationgpu.apk

7.Hands 手部检测和跟踪
下载 handdetectiongpu.apk

 

8.Holistic 全面跟踪
下载 holistictrackinggpu.apk

9.Instant Motion Tracking 即时运动跟踪
下载 instantmotiontracking.apk

10.Iris 虹膜
下载 iristrackinggpu.apk

 

 

11.KNIFT (Template-based Feature Matching)
下载 templatematchingcpu.apk
MediaPipe KNIFT is a template-based feature matching solution using KNIFT (Keypoint Neural Invariant Feature Transform).

MediaPipe KNIFT是一种基于模板的特征匹配解决方案，使用KNIFT（Keypoint Neural Invariant feature Transform）。

 

 

12.Object Detection
下载 objectdetectiongpu.apk

 

 

13.Objectron (3D Object Detection) 三维目标检测
下载 Objectron (3D Object Detection) 三维目标检测.zip

 

 

 

 

14.Pose 姿势跟踪
下载 posetrackinggpu.apk

 

 

 

————————————————
版权声明：本文为CSDN博主「煞比猫」的原创文章，遵循CC 4.0 BY-SA版权协议，转载请附上原文出处链接及本声明。
原文链接：https://blog.csdn.net/chitiantong/article/details/114937760