likedan/awesome-coreml-models

Largest list of models for Core ML (for iOS 11+)

awesome-coreml-models's Issues

Food101 model

Sorry that I don't conform to your issue guide but my issue is just to improve an already existing model in your collection.

I saw that you added images with the input types and the output types of each model to your README and I therefore wanted to tell you that I finally converted the Food101 CoreML model in a matter where it now accepts CVPixelBuffer instead of MLMultiArray as an input.

Just a little update :)

Nudity model

License

BSD-2

Summary

This model detects nudity and it's score from an image. The original dataset is Yahoo's OpenNSFW.

Model URL

Since this is too big for Github I have a GoogleDrive link:
https://drive.google.com/open?id=0B5TjkH3njRqncDJpdDB1Tkl2S2s

Demo URL

https://github.com/ph1ps/Nudity-CoreML

Samples

Input: Output: SFW (100%), NSFW (0%)
Input: Image, Output: NSFW (80%), SFW(20%)

This gives you back both NSFW and SFW score in order to make people able to decide what their personal threshold is. Let's say there is a picture with NSFW - 70%, some people might consider this as safe but others not. Therefore they can say everything from 70% and down is SFW and everything from 71% to 100% is NSFW.

Checklist

• Only one item is in this issue
• The model info contains all the required fields
• The demo project is compilable
• Has proper reference
• If this model takes image as input, an image type is selected instead of multiarray
  (See https://developer.apple.com/wwdc17/710 if you don't know how to do this)

Traffic Sign Detection Benchmark

This model should detect and recognize live video traffic signs. If there is a speed limit fo example the app should detect and recognize this shield.

There is a lot of sources...

Model URL

http://btsd.ethz.ch/shareddata/
http://benchmark.ini.rub.de/?section=gtsdb&subsection=news

Demo URL

http://www.vision.ee.ethz.ch/~timofter/

New Model:NIMA

License

MIT

Summary

Automatically learned quality assessment for images has recently become a hot topic due to its usefulness in a wide variety of applications such as evaluating image capture pipelines, storage techniques and sharing media. Despite the subjective nature of this problem, most existing methods only predict the mean opinion score provided by datasets such as AVA [1] and TID2013 [2]. Our approach differs from others in that we predict the distribution of human opinion scores using a convolutional neural network. Our architecture also has the advantage of being significantly simpler than other methods with comparable performance. Our proposed approach relies on the success (and retraining) of proven, state-of-the-art deep object recognition networks. Our resulting network can be used to not only score images reliably and with high correlation to human perception, but also to assist with adaptation and optimization of photo editing/enhancement algorithms in a photographic pipeline. All this is done without need for a "golden" reference image, consequently allowing for single-image, semantic- and perceptually-aware, no-reference quality assessment.

Model URL

Model Link

Demo URL

PhotoAssessment

Samples

Input: Image[http://i1.hdslb.com/bfs/archive/9892e6f032425fc9e9831fa1ed855318c12702ad.jpg], Output: Double 10 vector [1.614268398952845e-06,0.0006238522473722696,0.02812075242400169,0.2070262581110001,0.3266536593437195,0.332121878862381,0.09285952150821686,0.01259173825383186,2.718873588491988e-07,4.697789393048879e-07]

Checklist

• Only one item is in this issue
• The model info contains all the required fields
• The demo project is compilable
• Has proper reference
• If this model takes image as input, an image type is selected instead of multiarray

New Model: Fast-Style-Transfer

License

My Code is MIT, though the model requires an agreement for commercial use

Summary

I implemented Fast Style Transfer in CoreML. This is similar to Fast Neural Style, but is TensorFlow.
My article explains the process.

Model URL

https://github.com/lengstrom/fast-style-transfer

Demo URL

https://medium.com/@rambossa/diy-prisma-fast-style-transfer-app-with-coreml-and-tensorflow-817c3b90dacd
https://github.com/mdramos/fast-style-transfer-coreml

Samples

Here are some CoreML models I created in the process: https://drive.google.com/drive/folders/1CBSanBHbXC5-bJNTTk3-r1WSq56z0eKG?usp=sharing
-- just include these in the ios app before running

Checklist

• Only one item is in this issue
• The model info contains all the required fields
• The demo project is compilable (AFTER INCLUDE THE MLMODELS)
• [?] Has proper reference
• If this model takes image as input, an image type is selected instead of multiarray
  (See https://developer.apple.com/wwdc17/710 if you don't know how to do this)

fyi - GAN

https://github.com/zedge/DynamicCoreML

Cool Start Up

This is not an issue I just wanted to make sure this repo had seen this cool start up:

https://lobe.ai/examples

They claim to have lots of CoreML Models from that example page. No way to download from what I saw though.

Regardless, I signed up for their beta.

All links leads to dead website coreml.store

https://downforeveryoneorjustme.com/coreml.store
Is this project still alive?

New Model: DocumentClassification

This model is also included in an open-source framework that can be used for document classification.

License

MIT

Summary

• Classifies documents into one of five categories (Business, Entertainment, Politics, Sports, Technology)
• Trained with 1,500 articles from the BBC. See "References" in README

Model URL

Model Link

Demo URL

Demo app is the NewsClassifier iOS app here

Samples

Screenshot from the demo app here

Checklist

• Only one item is in this issue
• The model info contains all the required fields
• The demo project is compilable
• Has proper reference
• If this model takes image as input, an image type is selected instead of multiarray
  (See https://developer.apple.com/wwdc17/710 if you don't know how to do this)

NamesDT

(not the author, just stumbled across this and seems it would fit in Text Analysis here)

License

MIT

Summary

A Demo application using CoreML framework for predicting gender from first names.
See Is it a boy or a girl? An introduction to Machine Learning

Model URL

CoreML model was converted from Scikit-learn Pipeline using coremltools python package.

Demo URL

NamesCoreMLDemo

Samples

In demo README

Checklist

• Only one item is in this issue
• The model info contains all the required fields
• The demo project is compilable
• Has proper reference

issue

Validate Links

Hello, I wrote a tool that can validate links in a README (or any file). It can be run when someone submits a pull request or pushes a commit to Awesome-CoreML-Models.

This tool is currently being used by

• https://github.com/enaqx/awesome-react
• https://github.com/matteocrippa/awesome-swift
• and many more

More examples

• https://travis-ci.org/matteocrippa/awesome-swift/builds/96526196 ok ✅
• https://travis-ci.org/matteocrippa/awesome-swift/builds/96722421 link redirected / rename 🔴
• https://travis-ci.org/dkhamsing/open-source-ios-apps/builds/96763135 bad link / project deleted 🔴
• https://travis-ci.org/dkhamsing/open-source-ios-apps/builds/95754715 dupe 🔴

If you are interested, connect this repo to https://travis-ci.org/ and add a .travis.yml file to the project (you can also use CircleCI or other CI tools).

See https://github.com/dkhamsing/awesome_bot for options, more information 😄

Size of CoreML model

Hi, I am trying to use CNNEmotions, but his size is too big.

Would it be possible to reduce its size?

Thanks!

Anyone had any success porting a GPT-3 model?

I'm in the middle of trying to figure out how to map up the transformers and such. But not very experienced and running into headaches. Anyone want to tag team?

Is there any font detection model?

Please fix AgeNet link

Places CNN downloads MobileNet

https://github.com/hollance/MobileNet-CoreML/raw/master/MobileNet.mlmodel

This doesn't seem to be the correct model linked up

Is there any traffic light (red, green, yellow) model?

Just need a model that can detect traffic lights and give me the color state of it. Thanks.

"TextRecognition" is not a CoreML example

TextRecognition - Recognizing text using ML Kit built-in model in real-time.

Links to project that uses ML Kit (Google Firebase) and not CoreML (Apple) framework and model.

How to convert Tensorflow models to CoreML model?

Models for exercising

Do you have any mlmodels using activity classification for exercising? Like push ups and jumping jacks?

Not able to open links

most of the links in readme are not opening

New Model: PoseEstimation

License

Apache License 2.0

Summary

edvardHua implements PoseEstimationForMobile estimating human pose from a picture for mobile. And I make demo for that on iOS.

• Estimate human pose
• Train AI Challenger dataset that is single person image dataset contain about 20,000

Model URL

Model URL

Demo URL

Demo URL

Samples

Input

Image[URL]

Output

Heatmap[Array<Array<Array>>]

model	cpm	hourglass
output	[96, 96, 14]	[48, 48, 14]

Checklist

• Only one item is in this issue
• The model info contains all the required fields
• The demo project is compilable
• Has proper reference
• If this model takes image as input, an image type is selected instead of multiarray
  (See https://developer.apple.com/wwdc17/710 if you don't know how to do this)

I don't see any model for lowercase/uppercase text recognition (printed text... not handwritten) and digits.

License

Summary

Model URL

Demo URL

Samples

Checklist

• Only one item is in this issue
• The model info contains all the required fields
• The demo project is compilable
• Has proper reference
• If this model takes image as input, an image type is selected instead of multiarray
  (See https://developer.apple.com/wwdc17/710 if you don't know how to do this)