一年前,我使用OpenCV和Mediapipe在Python中完成了很多姿势估计。最近我读了Tensorflow博客(https://blog.tensorflow.org/2021/08/3d-pose-detection-with-mediapipe-blazepose-ghum-tfjs.html),关于可能实现三维姿势估计的内容。这引起了我的兴趣,所以我学了一点JavaScript开始我的旅程。经过3天的尝试,我意识到我的代码不起作用,JavaScript没有报错,但什么都没发生。我与一个已经使用JavaScript一个月的朋友密切合作,但他无法帮助我,因为他从未涉及过AI。我从未接触过它,所以我不知道在JS中await如何工作。这可能是个问题呢。
<body>
<div id="app"></div>
<video id="video" controls><source src="vid.mp4" type="video/mp4"/></video>
<script src="https://cdn.jsdelivr.net/npm/@tensorflow-models/pose-detection"></script>
<script src="https://cdn.jsdelivr.net/npm/@tensorflow/tfjs-core"></script>
<script src="https://cdn.jsdelivr.net/npm/@tensorflow/tfjs-converter"></script>
<script src="https://cdn.jsdelivr.net/npm/@tensorflow/tfjs-backend-webgl"></script>
<script src="https://cdn.jsdelivr.net/npm/@mediapipe/pose"></script>
<script>
const model = poseDetection.SupportedModels.BlazePose;
const detectorConfig = {
runtime: 'tfjs',
modelType: 'full'
};
const detector = await poseDetection.createDetector(model, detectorConfig);
const video = document.getElementById('video');
const poses = await detector.estimatePoses(video);
document.getElementById("app").innerHTML = str(poses[0].keypoints3D);
</script>
</body>
我的目标是将检测到的姿势数组“打印”到屏幕上,以便我可以看到它是否起作用。有什么想法吗?
import * as poseDetection from '@tensorflow-models/pose-detection';
import '@mediapipe/pose';
async function estimatePosesOfVideo(videoelement) {
const model = poseDetection.SupportedModels.BlazePose;
const detectorConfig = {runtime:'mediapipe',modelType:'full'};
const detector = await poseDetection.createDetector(model, detectorConfig);
const poses = await detector.estimatePoses(videoelement);
return poses
}
const videoelement = document.getElementById('video');
const poses = estimatePosesOfVideo(videoelement);
console.log(poses)
我对这些完全没有头绪。