Image Prediction with Tensorflow JS on simple REACT App | Google's Teachable Machine

Image Prediction with Tensorflow JS on simple REACT App | Google's Teachable Machine
machine-learning tutorial, machine learning basics, teachable machine tutorial, teachable machine google, teachable machine tutorial 1, teachable machine tutorial 2, teachable machine tutorial 3, image prediction machine learning, describe image prediction 2021
The project can be found in my GitHub: github.com/dev...
+++++ IMAGE FROM LOCAL STORAGE +++++++
If you are looking to try an image from your local system instead of an image URL, you can clone the branch " img-from-local-storage" and it will simple example App with EJS and Node JS
git clone --branch img-from-local-storage github.com/dev... and follow README to execute the application.
Thanks for watching, sharing your feedback, and requesting to subscribe to my channel if you haven't done yet as this really keeps motivating me to continue. Thanks!

[@akshhay]

Thank you sir, this was what i was searching exactly for

[@hafizhzulian_]

how to predict image that we upload, can you help me ?

[@farouktouil5036]

put the app on github please

[@somnathpaul2020]

I know its quite a bit late, but here you go: github.com/developedbysom/Techablemachine.git

[@AarushKulkarni]

@@somnathpaul2020 The github does not have the actual code

[@somnathpaul2020]

@@AarushKulkarni You need to clone the application from master brunch. , please use the following command to clone this app: git clone github.com/developedbysom/Techablemachine.git --branch master.

[@AarushKulkarni]

@@somnathpaul2020 Thanks

[@gigachad3861]

bro I have created a model with teachable machine and I'm using Tensorflow js to get predictions it works but its really slow because of model.predict.dataSycn() function takes time to return predictions .
my code is as follows -:
//
const handleCameraStream = (imageAsTensors) => {
try {
} catch (e) {
// console.log("Tensor 1 not found!");
}
const loop = async () => {
// && detected == true
if (model !== null) {
if (frameCount % makePredictionsEveryNFrames === 0) {
const imageTensor = imageAsTensors.next().value;
await getPrediction(imageTensor);
// .catch(e => console.log(e));
}
}
frameCount += 1;
frameCount = frameCount % makePredictionsEveryNFrames;
requestAnimationFrameId = requestAnimationFrame(loop);
};
loop();
//loop infinitely to constantly make predictions
};
//
const getPrediction = async (tensor) => {
// if (!videoLink) {
if (!tensor) {
console.log("Tensor not found!");
return;
}
//
const imageData2 = tensor.resizeBilinear([224, 224]);
// tf.image.resizeBilinear(tensor, [224, 224]);
const normalized = imageData2.cast("float32").div(127.5).sub(1);
const final = tf.expandDims(normalized, 0);
//
console.time();
const prediction = model.predict(final).dataSync();
console.timeEnd();
console.log("Predictions:", prediction);
}
Please help brother my camera is physically lagging when I use the following . I calculated the time it takes to get results from the model.predict.dataSync() function it takes approx. 500ms to get predictions from it . I want this entire process to be instant so that my camera doesn't lag . pls help bro its urgent