Skip to content

Conversation

@abdou31
Copy link

@abdou31 abdou31 commented Jul 9, 2019

Hello,
Can you add please an example for face detection using Tensorflow lite on Android device?
Thanks

copybara-github and others added 30 commits April 11, 2019 09:50
PiperOrigin-RevId: 243091045
PiperOrigin-RevId: 243358359
PiperOrigin-RevId: 243883601
PiperOrigin-RevId: 243910948
PiperOrigin-RevId: 243917454
PiperOrigin-RevId: 243920143
Also add UI to control GPU/CPU execution, as well as model selection.

PiperOrigin-RevId: 244043941
PiperOrigin-RevId: 244117132
…0 and cl/244080246

PiperOrigin-RevId: 244290009
This has failed twice now, and both times unpinning fixed it. So I think the pinned version is causing problems instead of preventing them.

PiperOrigin-RevId: 244370763
PiperOrigin-RevId: 244418529
PiperOrigin-RevId: 245794809
PiperOrigin-RevId: 246071783
PiperOrigin-RevId: 246160583
This example has been migrated to the new TensorFlow Examples
repo @ https://github.com/tensorflow/examples/tree/master/lite.

RELNOTES=Removed TensorFlow Lite Android example (moved to new examples repo).
PiperOrigin-RevId: 246173733
…se the new `TensorFlowLiteSwift` CocoaPod. Objective-C developers should use the new `TensorFlowLiteObjC` CocoaPod.

PiperOrigin-RevId: 246365603
PiperOrigin-RevId: 246656018
PiperOrigin-RevId: 246666696
… entire loss by global batch size. Regularizer losses should not be scaled by batch size, only by number of replicas.

PiperOrigin-RevId: 246680480
…ions` struct. Error logging is enabled by default to match ObjC behavior.

PiperOrigin-RevId: 248392591
…ages_of_clothing-description

PiperOrigin-RevId: 248418742
yashk2810 and others added 16 commits June 11, 2019 14:54
PiperOrigin-RevId: 252704441
Conversion script fails for a model generated from latest tf.js
(1.1.2) with an error that reads:

"AttributeError: 'module' object has no attribute 'set_keras_submodules'"

Bumping Keras version to 2.2.4 appears to solve the problem. A more long-term solution would probably be to update the entire example to use TF 2.0.

PiperOrigin-RevId: 252851711
… new `TensorFlowLiteSwift` CocoaPod. Objective-C developers should use the new `TensorFlowLiteObjC` CocoaPod.

PiperOrigin-RevId: 253241005
PiperOrigin-RevId: 253320957
PiperOrigin-RevId: 253790284
PiperOrigin-RevId: 253883062
PiperOrigin-RevId: 254025612
The sample now uses the `TensorFlowLiteSwift` CocoaPod. Objective-C
developers should use the new `TensorFlowLiteObjC` CocoaPod.

PiperOrigin-RevId: 254031296
PiperOrigin-RevId: 254404628
PiperOrigin-RevId: 255446790
@review-notebook-app
Copy link

Check out this pull request on ReviewNB: https://app.reviewnb.com/tensorflow/examples/pull/73

You'll be able to see visual diffs and write comments on notebook cells. Powered by ReviewNB.

@googlebot
Copy link

Thanks for your pull request. It looks like this may be your first contribution to a Google open source project (if not, look below for help). Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

📝 Please visit https://cla.developers.google.com/ to sign.

Once you've signed (or fixed any issues), please reply here (e.g. I signed it!) and we'll verify it.


What to do if you already signed the CLA

Individual signers
Corporate signers

ℹ️ Googlers: Go here for more info.

@googlebot googlebot added the cla: no CLA has not been signed label Jul 9, 2019
@lamberta
Copy link
Member

lamberta commented Jul 9, 2019

Hi. Your pull request is not set up correctly and is merging from master into a patch branch.
Closing this but feel free to submit another, thanks.
(Also, please sign the CLA :)

@lamberta lamberta closed this Jul 9, 2019
@jdduke
Copy link
Member

jdduke commented Jul 12, 2019

Hi @abdou31, we generally don't accept wholesale example contributions without prior discussion, and it's unlikely that we'd accept a face-detection sample at this time. That said, if you want to host the sample elsewhere, we're considering building a link repository for samples that use TensorFlow Lite.

@abdou31
Copy link
Author

abdou31 commented Jul 12, 2019

Hello @jdduke, my request to a face-detection sample is that other samples (e.g object detection, image classification) can't help to solve the problem of extraction predicts from tflite model ( eye region detection with landmarks in my case ).
The first reason is that image classification and object detection used the back camera ( not the front camera ).
The second reason is that I don't have labels.txt (the file that contains classes names ) because of my project use regression not classification ( predict the eye region and draw landmarks ).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

cla: no CLA has not been signed

Projects

None yet

Development

Successfully merging this pull request may close these issues.