Stepwise implementation of Fruits Prediction Android App using Tensorflowlite|Teachable ML|Quantized

[@AghaKhan9]

What a great video and precise explanation with demo. I needed that information and yu made it very easy for me. Thank you so much brother...

[@dineshraturi]

Glad u liked it :)

[@akashde8919]

@@dineshraturi sir is python necessary for machine learning or we can do it on java

[@dineshraturi]

You can do it on any language. Math is necessary. But python and R provides lots of support library which you don’t need to write it manually. Thats they r most popular.

[@akashde8919]

@@dineshraturi sir one more question I have learnt java at medium and now I m persuing to advance but if I learn python can I apply python program in android studio and how ? PLEASE reply🙏

[@jasimchouhan4784]

Wow, before this I thought It is very hard to learn but you changed my perception. Please make more video on android+Nlp and ml

[@dineshraturi]

Thanks jasim.

[@vigneshkumar6451]

Wowww nice content. You can start teaching data science from scratch.

[@dineshraturi]

Thanks. Yes that’s the plan. Stay tuned

[@muhammadabdullahorakzai587]

great job kindly create more tutturials on this topic

[@dineshraturi]

Thank you ,Sure stay tuned.

[@rinkur3880]

Yes sir i also need more videos.

[@abineshrl204]

Keep uploading content like this in English

[@dineshraturi]

Sure buddy. Stay tuned. Idea is to education for all.

[@rudreshsp3236]

instead of importing project, u could have started new android project, that helps everyone a lot

[@dineshraturi]

Namesty Rudresh , sure , Yes its for absolute beginners. Next project i am preparing totally end to end. 👍 stay tuned and do subscribe to my channel it will help me to grow for sure.

[@sirireddy129]

This is so precise and to point.. was very helpful, thank you !

[@dineshraturi]

Please share and subscribe. New end to end project is coming soon.

[@srinivasachary7392]

Superb. Dinesh

[@pravindesai6687]

Hi dinesh , nice video explaination. The code ran successfully but classifier on mobile is not using the custom model but other default models . Can you explain why that might be the case?

[@cryptostar1427]

Thank you man ! please make us more videos on these ! I still did not understand the part where the voice is there and speaks for the classified image

[@arshassija]

Thanks for awesome tutorial.

[@AM-bk6te]

Sir I created this image classification app and it is running successfully. Sir can you please make a video on how to add the classified image's text to speech convertor? It will be a great help sir. I tried my best but not able to do it.

[@dineshraturi]

okay sure. I will make it soon within this week :)

[@dineshraturi]

Video is live - kzbin.info/www/bejne/qaang6hrf8-mbrc

[@AsadKhan-js5le]

Can you please provided source code for image classification using android app?

[@dineshraturi]

@@AsadKhan-js5le please watch this video for source code - kzbin.info/www/bejne/qaang6hrf8-mbrc

[@diyandassembling2907]

Great work. Keep posting.

[@azamsayeed3644]

Keep up the good work!!

[@jayashankar6436]

hai sir expecting more from u.plz do upload more vedioes on object detection machine learning Ai ect...

[@dineshraturi]

Sure i am dedicated towards it. Just wait for few days. Let me migrate my goods back to my hometown 🤟🙏

[@jayashankar6436]

@@dineshraturi yes sir we are waiting for u❤❤

[@ramavol]

Thank-U, hello from Russia!

[@shambhaviverma1353]

wonderful explanation.Thanks

[@zhitao7198]

Nice tutorial and thank you for sharing. Is it a series tutorial? Because I didn't saw you making Camera Activity. Because I plan to add this feature to my final year project.

[@dineshraturi]

U can use my code.

[@rinkur3880]

No its single video. Camera activity u will find in his code.

[@jopadjr]

Great video. Are there any updates on this? Thanks

[@rinkur3880]

Wow 🤩

[@vijendraelgroup4369]

Good work 👏👏

[@swamynathant7764]

really it's very useful to me

[@dineshraturi]

Thanks plz do subscribe to my channel it will help me to grow. 🤟

[@justcats2184]

Thank u bro this is very informative video, keep the good work

[@dineshraturi]

Many thanks

[@mishakhan7643]

Nicely explained 👍 BRO CAN U PLZ SHOW THE SAME MODULE BUT WITH PASSING PICTURE TAKEN FROM CAMERA (CAPTURE) AND AFTER PREDICTION SEND TO PARTICULAR FOLDER ( BANANA TO BANANA, APPLE TO APPLE FOLDER)

[@dineshraturi]

Yes it is possible. Misha as of now i have many other requests to full fill. Allow me sometime i will do that for you :) stay tune.

[@mishakhan7643]

@@dineshraturi thanks a lot for replying just guide me a little bit if possible as I don't have much Time :'(

[@dineshraturi]

Sure Why not. Join me over linkedin send me the details.

[@mishakhan7643]

@@dineshraturi I am not on linkdyn :'(, actually I am a student and working on my project in which I have to pass pic and through ML pic after detection send to particular folder ( T-shirt into top folder) can u plz show me the reference android code or make a short video I will give u a short out (will like and comment on all your videos as I already subscribed 😁) u can add code in the same fruits video just capture pic and send them to banana folder etc plzzzzzzzz😖

[@mishakhan7643]

@@dineshraturi plz help kr den mujhy submit krwana h :'((

[@destinyjames6117]

Hi kind sir I cant find the "tflite" folder which contains "ClassifierFloatMobileNet.java" to edit the "protected String getModalPath()" as mentioned at 11:03 Even at tensorflow github link (github.com/tensorflow/examples/tree/master) I cant find the folder specified too This resulted in me using the model trained by them. Can sir please advise?

[@awkamweng]

Any help on this ?

[@awkamweng]

Found the code you can copy from this repo github.com/AndroidArena/CurrencyDetectorAndroid

[@destinyjames6117]

@@awkamweng Yes, google updated the file arrangements in 2020 i think. Go to libsupport -> src -> main -> java -> org.tensorflow.... And boom you're in. Also, the model is not at the placed specified in the video, it is in models -> src -> main -> assets Just that for now I am facing accuracy issues, even though i have high accuracy at the google trainable machine website, anyone else facing this issue..? Example, I am training differentiation between hand gesture (good and bad), and it keeps showing bad for some reason.

[@dineshraturi]

Guys u can use my currency detector code from github link already mentioned above by weng.

[@dineshraturi]

For accuracy issue --> you have to hypertune some parameters and increase ur training data.

[@jyotsna381]

Amazing

[@xl0pate0lx]

How can I save the result in a variable? I would like to classify clothes with pictures and safe the result of the prediction in a variable.

[@dineshraturi]

Use shared preferences in android to store the data

[@DeepakKumar-Dalsinghsarai]

Bro, I follow this tutorial step by step app run successfully, but don't recognize any materials.

[@dineshraturi]

Ok no worries. It means problem is with training. Try to train again and don’t forget to use correct model name. For floating and quantized.

[@amuchand47]

@@dineshraturi I have trained it again , but nothing is showing there. Please do help me.

[@DeepakKumar-Dalsinghsarai]

@@amuchand47 Change default Package Name & thank me later 👍

[@mohitsesharatan]

@@DeepakKumar-Dalsinghsarai Could you elaborate on changing default package name as I've been facing the same issue

[@dineshraturi]

Hi mohit follow this. - stackoverflow.com/questions/16804093/rename-package-in-android-studio

[@babafakruddin5399]

Good job

[@dineshraturi]

Thank you baba Ji 🤟

[@heetpatel3997]

Hello Sir.. I've a problem while converting quantized model . i've tried for several times but got the same result

[@dineshraturi]

Yes issue is from teachable ml website. Keep trying.

[@sajaata2006]

How to add audio to read the result when the result appears?

[@kartikeychauhan2980]

Wowwww

[@mishakhan7643]

BRO I M WORKING ON A FASHION APP, WHICH IS ABLE TO DETECT TYPE OF CLOTHES( TOP, TROUSER, SHOES ETC) BUT NEED TO PASS PIC TAKEN FROM CAMERA AND AFTER PREDICTION SAVED IN FOLDERS PLZZZ HELPPPP

[@AsadKhan-js5le]

Misha have you developed an android app ? i need help for app development

[@ikelloydyana4275]

Does tensorflow detects such as nudity, weaponry or any offensive images?

[@freelancedeveloper2591]

I want to create a an app for blind people for counting money..so the app should be able to recognise each money note or coin and give a voice commands about the count ..how can i achieve this..

[@dineshraturi]

Its very easy brother. Likewise fruit classifier just make a note classifier model and upon successful predictions on App add a music file there in Android.

[@chathuryaatthanayake5940]

Thank you so much

[@ranniesumacot4576]

Thanks. Do you have an updated tutorial video? I seem cannot follow the steps since that there are folder that are missing in github (i.e. Assets Folder, tflite folder).

[@dineshraturi]

I need to make one, let me check

[@dionrizqi]

What algorithm did you use? Rcnn?

[@dineshraturi]

CNN - Resnet50 architecture

[@dionrizqi]

Brother, why are you not using box detector? Can you tell me how tow use the detector instead of using label?

[@dineshraturi]

Here is my tutorial- kzbin.info/www/bejne/jYW7eJyte7FogqM

[@shenglilim5485]

Hi Dinesh , I have to change the model to my model and replace to your model file. But was nothing happend and the percent not show when I detect something. Can you help me check something happend ?

[@ronniepelayo9639]

It's not working on my project, it was detectng other object

[@dikiwahyudi219]

I don't know what's missing, I replaced the teachable file, but when I run it, no data appears

[@parasmanisharma8820]

Great👍

[@alizeshan872]

if i want to only detect camera using custom model how can i do that ,...i have add multiple camera photo then trained them and add the model to my assets but it does detec t the camera i added earlier

[@shorouq6548]

It is an amazing and helpful video, could you please help me? I have to create an android application witg a feature from the AI, i want to add an easy and simple feature and don’t know what is it. I need a help if you can please

[@hrizvi8687]

Hello, can you please send me the code to detect the disease in leave. I have a model but i didnt understand how to deploy in android studio

[@pranaykumarpokala]

can you show us to create an app without USB connectivity? and also how to create a button to upload photos

[@dineshraturi]

USB connectivity is the base of Android dev. Everyone in the org. Follows the same. But yes through WIFI you can achieve the same. So you just need to goto mobile setting -> about -> click 10 times on build no. -->> you are a developer now. Now find Developer option and turn on USB debugging option thats it. You will be able to use usb now.

[@dineshraturi]

Also for button where you want to upload it ?

[@artsie6146]

my quantized folder keeps saying it's damaged or in an unknown format, any clue?

[@Zeeshan_Ahmed5588]

also facing this issue how you handle this ?

[@mohamedabdelgawad9965]

Hello, thanks for the great video but the app always gives the default layout that not reflecting my labels

[@dineshraturi]

Thanks mohamed. Okay it means your inferences are not working properly. Can you please check my another video of currency detection where i have given the full source code with models. Then again try this with that code only.

[@ParthPatel-gq5mg]

Sir help me , I convert this code into Koltin but it's not work ,camera not open ,

[@lofiworld5097]

Can you make an app using summarise model with redis cache...🙏🙏🙏

[@healersage2281]

sir what if i have a lot of inferences in my own customized TFLITE model (I didn't use teachable machine)? The application doesn't seem to run when using the model I made I think its because Im using a customized one? here is what the code looks like after I imported it to my project. try { Tflite model = Tflite.newInstance(context); // Creates inputs for reference. TensorBuffer inputFeature0 = TensorBuffer.createFixedSize(new int[]{1, 320, 320, 3}, DataType.FLOAT32); inputFeature0.loadBuffer(byteBuffer); // Runs model inference and gets result. Tflite.Outputs outputs = model.process(inputFeature0); TensorBuffer outputFeature0 = outputs.getOutputFeature0AsTensorBuffer(); TensorBuffer outputFeature1 = outputs.getOutputFeature1AsTensorBuffer(); TensorBuffer outputFeature2 = outputs.getOutputFeature2AsTensorBuffer(); TensorBuffer outputFeature3 = outputs.getOutputFeature3AsTensorBuffer(); // Releases model resources if no longer used. model.close(); } catch (IOException e) { // TODO Handle the exception } Can you make a vid that can solve this problem? It will be appreciated if you make one tutorial regarding this problem.

[@imskumarj]

In my chrome, the quantized model of my project isn't downloading... Please help sir. After a long time of processing it responds with, 'Something Went Wrong While Converting...'

[@dineshraturi]

Yes, i have reported this issue at google. But they said clear your cache and do not change tabs. Also please check a proper internet connection. Try to change the browser also if possible.

[@dineshraturi]

Please consider subscribing to my channel if it helps you. :)

[@rahulsh5237]

@@dineshraturi Same issue for me,after clearing cache also

[@dineshraturi]

@sudanshu @rahul [e-mail] send me ur dataset with 10-10-10 images. I will try at my end. And will send you the model.

[@lasindadilshan7860]

i want to get description of the high probablity fruit. how i do it?

[@dineshraturi]

There is percentage showing there on the basis of that whatever has >50% prob. Show the description in the Textview or in the dialogbox from local strings.xml file or may br from sqlite database.

[@chitrachitra1273]

In my android studio , classifier float efficient net file is missing,what is the problem? And mistakes I had done....I don't know...help me

[@dineshraturi]

Hi Chitra, use this code github.com/AndroidArena/CurrencyDetectorAndroid change the lables.txt and model. Watch my currency detection video for code detail.

[@chitrachitra1273]

@@dineshraturi ok bro... thank you

[@vinayreddy8683]

Is it same procedure if we have developed our own model on Jupiter notebooks?

[@dineshraturi]

Yes. Make sure your final model is in quantised tflite Format. I will come up with the tutorial soon. Stay tuned. Plz do subscribe.

[@vinayreddy8683]

@@dineshraturi I rarely subscribe on KZbin and subscribed to your channel. Your content is amazing and more practical oriented. Waiting for the video!!

[@dineshraturi]

Thanks vinay means alot. Plan is to balance between theory + end to end practical implementation.

[@dayojaiyed7721]

Awesome

[@syedshoaibabbas]

Sir ,when I build project t tflite pkg folder not created how we can do ??....

[@dineshraturi]

Goto my github and download currency detection projection. U will find all the files

[@hazkimshazkim1041]

Hallo sir, im a final year student and doing a project about plant identification using apps. In your opinion, can i use your method on doing machine learning or i should make Convolution neural network that is really hard for me to make it?

[@dineshraturi]

For major project use CNN manually. Don’t use this method.

[@hazkimshazkim1041]

@@dineshraturi alright sir thanks

[@ayeshamushtaq604]

@@hazkimshazkim1041 hi do you have that plant identification project, actually i am doing the same fyp project on plant disease identification

[@trishantsaxenaco]

There's no tflite folder in my android folder. What should I do?

[@dineshraturi]

Namesty bhai , You have to manually create it. Do one thing search my another video in this playlist currency detection app. Use that code. It will work. You just need to replace your tflite model with mine.

[@RiteshSingh-nv8cv]

you are correct , I have faced the same issue .The solution to this problem is you can download the source code of " currency detection app " from this link - " github.com/AndroidArena/CurrencyDetectorAndroid " and then just delete all file from " assets folder " then copy past your trained files i.e 1. labels.txt , 2. model.tflite , 3. model_unquant.tflite , over their then it will work .

[@dineshraturi]

Thanks bro. I have also updated my video description with this information. Stay tuned. Jald vapas aaunga 🆘 k baad. !!

[@anildogne3839]

Plz make one more video for vs code

[@trishantsaxenaco]

Did you make a video on how to convert Tensorflow model to Tensorflow Lite Model?

[@dineshraturi]

Not yet

[@sairamadithya9650]

Can you show the mobile part

[@hananmohmed4310]

i cant find assets folder what can i do plz?

[@rozhinrafiq9163]

Hi why in downlod file zip in converted_tflite_quantized to foder? error

[@dineshraturi]

Error is from teachable website. Many people are getting it.

[@kingstonsotelo6161]

please help. you have 2 examples folder: examples and examples-master. i only have examples-master. where is the other one? there is no assets folder in the masters

[@dineshraturi]

Use my github code which is mentioned in the another video of currency detection. Follow it.

[@ex_09_vikrantneve55]

There is no asset folder in my code. Will it work if i create manually

[@dineshraturi]

You can use my another code of currency detection from github.

[@ex_09_vikrantneve55]

@@dineshraturi use that but to use only two class what changes i have to do

[@dineshraturi]

Some intermittent issue is going on in teachable ml website. Hack - Make 3 categories and name it “others”

[@ex_09_vikrantneve55]

@@dineshraturi ok

[@mageduu]

it wont show the tflite where the 2 classifier at. how to fix it??

[@dineshraturi]

This is know Issue. Even i m also not able to resolve this.

[@mohammadridhopardomuan2486]

i cant find assests folder, help

[@jayaraghavendra9025]

hi Dinesh , i'm learning from your video but I didn't find assets folder in it?

[@dineshraturi]

Kindly find my second video of currency note detector in which i have mentioned the code. Use that one. Replace tflite accordingly.

[@bhaskargethe1811]

Can you do one project on retail if possible

[@dineshraturi]

Sure. Stay tuned. As of now I am making python for absolute beginners post that simplified applied statistics are planned. I will definitely try to make one. Kindly subscribe if you like my content. I m working hard to make it better day by day.

[@abineshrl204]

How much time its take to get trained...

[@dineshraturi]

Depending on your data size. As its a automated system so it is very quick. You can watch the whole video i have shown all the details there. Do let me know in case of any query.

[@chinmayl444]

quantized model is damaged, how to proceed?

[@PD-iu2sr]

Brother I m getting an error It shows Gradle sync failed SdkPlatformNotFoundException

[@dineshraturi]

Which code u r using?

[@PD-iu2sr]

@@dineshraturi after Teachable I have downloaded zip file than I open it into Androidstudio

[@dilkiabeysekera1830]

Sir, I have done everything in the video , but app getting close. what should I do

[@dineshraturi]

Can u plz share the android crash logs

[@Lila12386]

Same her e . Have you fix that problem

[@akmalyafi1470]

Do i have to install java jdk in order to run this on android studio?

[@dineshraturi]

Yes. It will come along with android studio.

[@bhupendranagda9563]

Does that already present 4 files in assert means that we are training on already present mobiletnet model??

[@dineshraturi]

Android internally use your tflite as mobilevnet. Try to delete all other files just put 2 tflite and labels.txt. Then run your app and recheck assest folder you will see mobilevnet is generated Do watch my new currency detector app video which is with voice.

[@bhupendranagda9563]

@@dineshraturi okay thanks

[@supratiksinha3885]

Can we apply the same concepts for videos as well ?

[@dineshraturi]

You need to create custom model in that case. And have to provide frame of a video at a time.

[@shubhamcalla3929]

Can I get document and srs .?

[@azullyaidris2074]

hi, is it possible to remove the tensorflow logo in the application?

[@Isha-ot6tc]

Is there a way, to download the application as an apk from Android studio, and then run on the mobile phone?

[@dineshraturi]

Goto build -> build apk -> now open project and find build folder inside it output then apk and find debug.apk. Now put that inside phone internal storage and install it manually. Thanks Kindly subscribe

[@bernardolaborde5310]

I have been facing some issues with the audio your audio seems to work but when I try to put the one I created it doesnt any suggestions?

[@dineshraturi]

What is the extension of that audio file ?

[@bernardolaborde5310]

@@dineshraturi Also while your project says this: "I/MediaPlayerNative: Pid:27408 MediaPlayer::start D/MediaPlayerNative: Action:start, CurrentState:MEDIA_PLAYER_STARTED" mine says this: "W/MediaPlayer-JNI: MediaPlayer finalized without being released "

[@Chhiling08]

how can i get data model for high accuracy to train my model is there any website I can download for data set without making myself

[@dineshraturi]

In Machine learning there is a concept of hyper parameters tuning. Through which you will be able to get good accuracy. 2. By adding more training data u can get more accuracy. 3. There are free dataset providers like Kaggle Google datasets Uci Data.gov etc Please subscribe if it helps you :)

[@chasingdreams8764]

Bro,.Can you make a video on how to add voice to this app?

[@dineshraturi]

Here is the video enjoy - kzbin.info/www/bejne/qaang6hrf8-mbrc For more on demand videos Please subscribe to my channel :)

[@mohaliyet]

Great. Could you create the same for flutter?

[@dineshraturi]

Didn’t explored the flutter for ML yet. Need to check

[@bhoomiailani7350]

I need help I am facing a issue.I have followed all the steps the app is also installed but there is no labels or names of the object.

[@dineshraturi]

Hi , do one thing follow my another video “currency note detector with voice” and use its code. Link mentioned in the description of that video. That code is working code you just need to replace your tflite model. Also that code has voice output also. Do watch that. Thanks.

[@bhoomiailani7350]

@@dineshraturiThanks will surely try it

[@gugasevero76]

Hi Dinesh, I don't know what happen with the Android Github Code, but inside the android project there is no tflite folder. So what i have to do? Since now, thank you for your help.

[@dineshraturi]

Namesty Gustavo , do one thing follow my another tutorial currency detector and use that code. You Just need to replace tflite code and thats it.

[@dineshraturi]

Link - kzbin.info/www/bejne/qaang6hrf8-mbrc

[@onlineaccadmy9659]

text recognition from image in python using machine learning .....plz iska bta dy

[@dineshraturi]

here is the video - Kindly subscribe kzbin.info/www/bejne/bHeqeoOmq751kNk

[@ganesharavindh8288]

I am trying to build drowsy driver detection android application. Please could you post a video on how to build it?

[@dineshraturi]

Great. Its very easy First option - if u r working on js model for it then Make an app with web browser and load that model on web within android without internally integrating tflite model. Second option - need to integrate tflite model in android. (You need some r&d )

[@nahombelete2160]

Hello brother! I follow every step but there is no tflite package in the android studio how can i fix that? Would you help me please?

[@dineshraturi]

kzbin.info/www/bejne/qaang6hrf8-mbrc watch this and use its code.

[@RiteshSingh-nv8cv]

you are correct , I have faced the same issue .The solution to this problem is you can download the source code of " currency detection app " from this link - " github.com/AndroidArena/CurrencyDetectorAndroid " and then just delete all file from " assets folder " then copy past your trained files i.e 1. labels.txt , 2. model.tflite , 3. model_unquant.tflite , over their then it will work .

[@49_tirtharajhalder49]

Bro i have only download the example master folder but i cant find the example folder

[@dineshraturi]

check my description for new code. it is 100% working

[@johndrahcir2641]

How will i add unknown? For example the image i analyzed is not on my classes

[@dineshraturi]

Use if else dude :)

[@johndrahcir2641]

@@dineshraturi thank you ill try that. how about ill use a different model. NOT the model i made from teachable machine? will it work?

[@johndrahcir2641]

@@dineshraturi if ill use if else. which part of the code will i use it?

[@ajinkyapatharkar9677]

On teachable machine site, I can only download tensorflow.js model but when I try to download tensorflow lite model, it says "something went wrong while converting" everytime. Can you help?

[@dineshraturi]

Its a intermittent issue from google. Try to download after 2-3 hrs.

[@reubens5693]

You might need to increase the number of classes, I faced the same issue while using just two classes. increase number of classes up to 4 and that solved my issue

[@dineshraturi]

Even 3 classes will also work.

[@harini7481]

I can't download the quantized model, I tried a lot of times.. Need help The error is something went wrong while converting

[@dineshraturi]

This is the intermittent issue from teachable ML. Try to clear cache and turn off the firewall and then restart your pc. Hope it will help. Plz report this issue to teachable ml who know cz of u they will fix some bug. Earlier one of the subscriber is also facing the same and his dataset was working at my pc and able to download.

[@harini7481]

@@dineshraturi ok I'll try that.. Thanks for the fast reply.. Subscribed!

[@ankitahegade7087]

Sir I have downloaded the android folder but in that assets folder is not there?

[@dineshraturi]

You hv to manually create the assest folder. Or visit my second video currency note detector and use that code. You just need to replace the tflite file in that code.

[@dineshraturi]

Link - kzbin.info/www/bejne/qaang6hrf8-mbrc

[@ankitahegade7087]

@@dineshraturi sir after that I created it manually but then in android studio the tflite codes are not there

[@dineshraturi]

You have to manually put it there.

[@anishthakur8938]

How can i download the actual neural network machine learning underhood code from teachable?

[@dineshraturi]

Under the hood for classification it uses Resnet50 architecture. so you can go to tensorflow github and download raw code for resnet50 and do the modification according to your need. After that train your model on TPU of google colab

[@coldzeragod4333]

DInesh Bhai there is no asset folder in my example-master-

[@dineshraturi]

Follow this - use this code - kzbin.info/www/bejne/qaang6hrf8-mbrc

[@yafercorralbarrera9089]

Hello Dinesh I am doing it but it takes so long to convert the quantized model is this normal? 😬

[@dineshraturi]

Yes

[@yafercorralbarrera9089]

Okay thanks Dinesh!! I wait like 6 hours but finally it says that what imposible to convert the quantized model What shoul I do? ☹️

[@dineshraturi]

How big is the data set ? Can u plz send me over drive if it is small ?

[@cellus5594]

can i do the training with my own model architecture? like using the VGG-16 convnet

[@dineshraturi]

Yes why not. After training convert them tflite with label.txt.

[@dineshraturi]

Very soon i am going to make project on same.

[@cellus5594]

@@dineshraturi hey can I reach you through e-mail, if you got spare time? I got some questions

[@dineshraturi]

Replied. Kindly check ! 🥂

[@cellus5594]

I re-reached you through e-mail. I have more questions that I hope you could help with, when and if you do have the time for. Thanks in advance! :)

[@sahoogoutam]

how to get tensorflow lite model from saved keras model?

[@dineshraturi]

You neee to convert it using tensorflow utility.