Deploy Ml model on android | android project java | android studio | cat dog app

this video is all about deploying your machine learning / deep learning model on the Android app using Java language.
by using this step by step guide you can deploy pretty much any deep learning model.
first step is to convert you are model to TFlite model.
second is to create basic gui.
third is to import the model and create backend functionality.
make sure you have latest version of Android studio installed.
if you want to follow along with this part you may use this model.
drive.google.c...
all the code is here
github.com/Paw...
THANKS FOR WATCHING...
LIKE + SHARE + SUBSCRIBE = support :)
contact me : 📞
facebook :
/ proogramminghub
twitter 🐦 :
/ programming_hut
github :
github.com/Paw...
discord :
/ discord
linkedin :
/ programminghut
youtube ❤️:
/ programminghutofficial
keys:
deploy ml model on Android,deep learning model on Android,machine learning Android app,deep learning Android app,how to deploy machine learning model,deployment of ml,machine learning model on Android

[@aguelejoseph5753]

YOU DID A REALLY REALLY GREAT JOB. Although i am doing something different and more complex but you have given me the headstart that I need. Thank you soo much.

[@yogietrypambudi6187]

wow what a good idea to using tensorflow for deep learning app android. but i"ve question when you click predict show the number accuracy data. it"s accurate?

[@programming_hut]

I don't really understood what you meant. I think you want to show the confidence of prediction right ? btw thanks for your compliments ...

[@greenheart3880]

explained it in really easy way THANKS 👍

[@paragbharadia2895]

Sir, at 0:46 which software/ide did you open?

[@programming_hut]

It’s sublime

[@ronixbhaskar8263]

Which software have been used ??

[@Sweety-rx8zq]

Why extension is as .h5? Any help would be appreciated

[@egamomusu]

Thank you..however the trained model is not available anymore in the given google drive link at the description section. Could you please upload the model again ? Thank you

[@matovushafik2372]

why is it that my app crashes whenever i click the predict button, anyways, great work, thank you

[@jeffr_ac]

Did It work?

[@lavanyal9418]

I also have same problem.Did you resolve your problem?If yes please let me know the solution

[@andrearomeroramos6586]

Hello!!! Could you pass me the code of your model please, the drive link does not work

[@tauseefrehman2]

Is this work for every kind of ML model?

[@shivamdubey4783]

sir do we need to learn java for this coz i only know python and tensorflow

[@programming_hut]

Yes basic understanding of java is required

[@swathinagiri3003]

I have used same program but instead of getting output the application is automatically closing

[@arushsingh5489]

Hey! How do I print my predicted label if I don't want to see numerical predictions?

[@ismail_..]

Hello sir,Can I use Kotlin language instead of java in tensorflow lite projects?

[@programming_hut]

Yes

[@__dekana__]

You are the best!!!!

[@gloriamasengo]

Hi can you show me how to display 5 classes not 2, the model is only showing me for 2

[@Sara-fp1zw]

that was nice thankyou!!!

[@m.abdullahkamran8805]

Hey Sir I am getting wrong output from tflite model when I converted it from keras model. Although output is perfectly fine when I use the same input in keras model.

[@AtticEdits]

Me too😢

[@fawniaavissa9711]

thankyou 8:35

[@kirtipatidar261]

hey can you tell me which linux-distro do you use

[@programming_hut]

Linux mint

[@localbus7008]

Sir i cant find modeltfLite file in your github

[@LAJianoran]

it shuts off when I use my model

[@localbus7008]

I can't understand the folders hierarchy..

[@amirayasmine8490]

thank you so much that was very helpful

[@programming_hut]

glad you liked it

[@seymatezcan4058]

Hello thank you for this video I have a question: How can I create model.tflite file on Jupyter notebook?

[@rushidesai2836]

Use Tensorflow with Python

[@programming_hut]

try this kzbin.info/www/bejne/fZaqnXqHbZWgkMk

[@rhk1991]

thank you kind sir, i really need this one.

[@Jerrod2444]

hi sir, how to get my output result if my result is String ?

[@nishilraiyarela5141]

Thanks it works

[@islembenamar1355]

how i can see the label for images ??

[@muhammadtalha2493]

hello sir, when i click the predict button application automatically close will u help ??

[@programming_hut]

what are inputs and outputs, if its image then what is input image size , if possible can you share code

[@muhammadtalha2493]

@@programming_hut i also train a model on googlecolab with cnn and then convert it in to tflite like the way You told and it happens the same when u clicked the predict button application shytdown

[@muhammadtalha2493]

image size is (200,200,3)

[@muhammadtalha2493]

@@programming_hut from tensorflow.keras.preprocessing.image import ImageDataGenerator from tensorflow.keras.preprocessing import image from tensorflow.keras.optimizers import RMSprop import matplotlib.pyplot as plt import tensorflow as tf from tensorflow import keras from keras.models import Sequential from tensorflow.keras import layers from keras.layers import Dense, Flatten, Conv2D, MaxPooling2D, Dropout #from keras.utils import to_categorical import numpy as np import os from keras import backend as K import cv2 from google.colab import drive drive.mount('/content/gdrive') cv2.imread("/content/gdrive/MyDrive/checkTest/training/diseased/0a01cc10-3892-4311-9c48-0ac6ab3c7c43___RS_GLSp 9352_270deg.JPG").shape train = ImageDataGenerator(rescale = 1/255) validation = ImageDataGenerator(rescale = 1/255) train_dataset = train.flow_from_directory( '/content/gdrive/MyDrive/checkTest/training', target_size=(200, 200), batch_size=3, class_mode = 'binary') validation_dataset = validation.flow_from_directory( '/content/gdrive/MyDrive/checkTest/validation', target_size=(200, 200), batch_size=3, class_mode = 'binary') train_dataset.class_indices model = tf.keras.models.Sequential([ tf.keras.layers.Conv2D(16,(3,3), activation='relu', input_shape = (200,200,3)), tf.keras.layers.MaxPool2D(2,2), # tf.keras.layers.Conv2D(32, (3,3), activation = 'relu'), tf.keras.layers.MaxPool2D(2,2), # tf.keras.layers.Conv2D(64, (3,3), activation = 'relu'), tf.keras.layers.MaxPool2D(2,2), # tf.keras.layers.Flatten(), # tf.keras.layers.Dense(512, activation = 'relu'), # tf.keras.layers.Dense(1, activation = 'sigmoid') ]) model.compile(loss = 'binary_crossentropy', optimizer = RMSprop(learning_rate = 0.001), metrics = ['accuracy']) model_fit = model.fit(train_dataset, steps_per_epoch = 3, epochs = 50, validation_data = validation_dataset) dir_path = '/content/gdrive/MyDrive/checkTest/testing' for i in os.listdir(dir_path): img = image.load_img(dir_path+'//'+i, target_size= (200,200)) plt.imshow(img) plt.show() X = image.img_to_array(img) X = np.expand_dims(X,axis = 0) images = np.vstack([X]) val = model.predict(images) if val == 0: print ("The leaf has disease") else: print ("The leaf is Healthy")

[@programming_hut]

so did you resized your bitmap accordingly in android studio

[@codewithchando971]

How I can Get the Confidence Value of The Prediction? Extra Code that's I need to add?

[@programming_hut]

the outputFeature variable contains the confidence of each class

[@m.abdullahkamran8805]

Sir how to normalize bitmap. My model takes pixel values from 0-1 but in android studio when i save image in bitmap it stores values from 0-255 in bitmap. Kindly tell how to covert bitmap values from 0-1.

[@Alex-bx6jf]

train the model without normalization or you can add pre-processing layers to the model itself during training

[@m.abdullahkamran8805]

@@Alex-bx6jf It is reducing overall accuracy of the model.

[@sarasaleh6751]

HI ,thank u for the this great video , but i can't upload my model it is bigger than 200 MB , is there is another way of upload or I should paste my model in assets .? thanks

[@programming_hut]

you can use api methods btw what your model do

[@sarasaleh6751]

@@programming_hut thank u , I don’t know how to use api methods .. my model it to recognize images type .. from an image

[@programming_hut]

may be this could help converting ml model to api : kzbin.info/www/bejne/bYXIaKWug5qeaMk how to make api android app : kzbin.info/www/bejne/n6GropyjbM2in9E

[@nitesharu1269]

can a pkl file be used in android studio

[@antapratama4735]

How to see name output with labels.txt and %predict?

[@safalghalegurung425]

hello sir im having trouble with model having multiple class can you make video on it as well?

[@safalghalegurung425]

there seems to be a video but its in kotlin and i am working on java project.

[@programming_hut]

are you looking for this ----> kzbin.info/www/bejne/bHbVk3mHl52EiZo

[@safalghalegurung425]

@@programming_hut well yes it is but i want it in java and this tutorial is based in kotlin.

[@arsalanahmed3488]

Can we use a teachable machine file here? using this method?

[@programming_hut]

as far as it is tflite it should work... I never tried but it should work

[@arsalanahmed3488]

@@programming_hut i trying to use it

[@programming_hut]

@@arsalanahmed3488 so did you got any error

[@arsalanahmed3488]

@@programming_hut i don't test it yet if it will be successfully run i will inform you

[@arsalanahmed3488]

@@programming_hut it's working proper but please let me know it's showing the result in number i don't want to show in number

[@Yussel_UG]

¡Gracias!

[@programming_hut]

Thanks 🙏

[@AshutoshKumar-yn6wf]

Hey bro can u tell me how to use full 1000 list of tensorflow in this code

[@saifchowdhury9728]

how can I add labels.txt code in this apps?

[@sivanit3150]

The video is too excellent!!!! But I'm getting error while using start activity for result, it says it is deprecated... What should i do now... Any suggestions... Also I'm getting an error as when i click on predict button, the app closes... I'm bit confused as to what would cause this. Thanks for your suggestion in advance!!

[@sureshxoela7273]

Thanks

[@handhikayp]

Sir, your drive links seems that you already remove the project into trash, can you renew the link? thanks

[@programming_hut]

I'll see

[@saipadmasree2048]

sir how can i create h5 file for my project?

[@deveshreekadu4087]

Plz upload TFLite file on github As it is not available on drive

[@programming_hut]

I'll see if i still have it on my laptop

[@robertgermia5758]

I used my model with leaf images for prediction. It successfully worked when I use leaf image. But it also gives prediction results for any image ex:-pencil. Can you please tell me how to validate only to predict the leaf images? I hope you will reply. Thank you. i think it is in if probability score > threshold value but I didn't have idead how to do it.

[@programming_hut]

when you trained your model you need add one more class called unknown and in that add whatever unknown images you can get . again train it and then try. I hope this could work.

[@dwiencabalquinto5545]

did you solve your problem?

[@davidcoyle958]

I assume this will work for any tflite model?

[@programming_hut]

yes as far as follow same steps

[@davidcoyle958]

@@programming_hut great thank u bro

[@saquelainmokhtar6016]

thanks

[@ywf98]

thanks