119 KiB
119 KiB
Awesome TensorFlow Lite !Awesome (https://awesome.re/badge.svg) (https://awesome.re) !PRs Welcome (https://img.shields.io/badge/PRs-welcome-brightgreen.svg?style=flat-square)
(http://makeapullrequest.com) !Twitter (https://img.shields.io/badge/Twitter-%40margaretmz-blue) (https://twitter.com/margaretmz)
TensorFlow Lite (https://www.tensorflow.org/lite) is a set of tools that help convert and optimize TensorFlow models to run on mobile and edge devices. It's currently running on more than 4
billion devices! With TensorFlow 2.x, you can train a model with tf.Keras, easily convert a model to .tflite and deploy it; or you can download a pretrained TensorFlow Lite model from the
model zoo.
This is an awesome list of TensorFlow Lite models with sample apps, helpful tools and learning resources -
⟡ Showcase what the community has built with TensorFlow Lite
⟡ Put all the samples side-by-side for easy reference
⟡ Share knowledge and learning resources
Please submit a PR if you would like to contribute and follow the guidelines here (CONTRIBUTING.md).
## Contents
- Past announcements: (#past-announcements)
- Models with samples (#models-with-samples)
- Computer vision (#computer-vision)
- **Classification** (#classification)
- **Detection** (#detection)
- **Segmentation** (#segmentation)
- **Style Transfer** (#style-transfer)
- **Generative** (#generative)
- **Post estimation** (#post-estimation)
- **Other** (#other)
- Text (#text)
- Speech (#speech)
- Recommendation (#recommendation)
- Game (#game)
- Model zoo (#model-zoo)
- TensorFlow Lite models (#tensorflow-lite-models)
- TensorFlow models (#tensorflow-models)
- Ideas and Inspiration (#ideas-and-inspiration)
- ML Kit examples (#ml-kit-examples)
- Plugins and SDKs (#plugins-and-sdks)
- Helpful links (#helpful-links)
- Learning resources (#learning-resources)
- Blog posts (#blog-posts)
- Books (#books)
- Videos (#videos)
- Podcasts (#podcasts)
- MOOCs (#moocs)
Past announcements:
Here are some past feature annoucements of TensorFlow Lite:
⟡ Announcement of the new converter (https://groups.google.com/a/tensorflow.org/d/msg/tflite/Z_h7706dt8Q/sNrjPj4yGgAJ) - MLIR
(https://medium.com/tensorflow/mlir-a-new-intermediate-representation-and-compiler-framework-beba999ed18d)-based and enables conversion of new classes of models such as Mask R-CNN and Mobile
BERT etc., supports functional control flow and better error handling during conversion. Enabled by default in the nightly builds\.
⟡ Android Support Library (https://github.com/tensorflow/tflite-support/tree/master/tensorflow_lite_support/java) - Makes mobile development easier (Android
(https://github.com/tensorflow/examples/blob/master/lite/examples/image_classification/android/EXPLORE_THE_CODE.md) sample code).
⟡ Model Maker (https://www.tensorflow.org/lite/guide/model_maker) - Create your custom image & text (https://github.com/tensorflow/examples/tree/master/tensorflow_examples/lite/model_maker)
classification models easily in a few lines of code. See below the Icon Classifier for a tutorial by the community.
⟡ On-device training (https://blog.tensorflow.org/2019/12/example-on-device-model-personalization.html) - It is finally here! Currently limited to transfer learning for image classification
only but it's a great start. See the official Android (https://github.com/tensorflow/examples/blob/master/lite/examples/model_personalization/README.md) sample code and another one from the
community (Blog (https://aqibsaeed.github.io/on-device-activity-recognition) | Android (https://github.com/aqibsaeed/on-device-activity-recognition)).
⟡ Hexagon delegate (https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/g3doc/performance/hexagon_delegate.md) - How to use the Hexagon Delegate to speed up model inference
on mobile and edge devices. Also see blog post Accelerating TensorFlow Lite on Qualcomm Hexagon DSPs (https://blog.tensorflow.org/2019/12/accelerating-tensorflow-lite-on-qualcomm.html).
⟡ Model Metadata (https://www.tensorflow.org/lite/convert/metadata) - Provides a standard for model descriptions which also enables Code Gen and Android Studio ML Model Binding
(https://www.tensorflow.org/lite/inference_with_metadata/codegen).
Models with samples
Here are the TensorFlow Lite models with app / device implementations, and references.
Note: pretrained TensorFlow Lite models from MediaPipe are included, which you can implement with or without MediaPipe.
Computer vision
Classification
│ Task │ Model │ App | Reference │ Source │
├─────────────┼──────────────────────────────────────────────┼────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┼───────────┤
│Classificatio│MobileNetV1 (download │Android (https://github.com/tensorflow/examples/tree/master/lite/examples/image_classification/android) | iOS │tensorflow.│
│n │(https://storage.googleapis.com/download.tenso│(https://github.com/tensorflow/examples/tree/master/lite/examples/image_classification/ios) | Raspberry Pi │org │
│ │rflow.org/models/tflite/mobilenet_v1_1.0_224_q│(https://github.com/tensorflow/examples/tree/master/lite/examples/image_classification/raspberry_pi) | Overview │ │
│ │uant_and_labels.zip)) │(https://www.tensorflow.org/lite/models/image_classification/overview) │ │
│Classificatio│MobileNetV2 │Recognize Flowers on Android Codelab │TensorFlow │
│n │ │(https://codelabs.developers.google.com/codelabs/recognize-flowers-with-tensorflow-on-android/#0) | Android │team │
│ │ │(https://github.com/tensorflow/examples/tree/master/lite/codelabs/flower_classification/android) │ │
│Classificatio│MobileNetV2 │Skin Lesion Detection Android (https://github.com/AakashKumarNain/skin_cancer_detection/tree/master/demo) │Community │
│n │ │ │ │
│Classificatio│MobileNetV2 │American Sign Language Detection | Colab Notebook │Community │
│n │ │(https://colab.research.google.com/drive/1xsunX7Qj_XWBZwcZLyjsKBg4RI0DNo2-?usp=sharing) | Android │ │
│ │ │(https://github.com/sayannath/American-Sign-Language-Detection) │ │
│Classificatio│CNN + Quantisation Aware Training │Stone Paper Scissor Detection Colab Notebook │Community │
│n │ │(https://colab.research.google.com/drive/1Wdso2N_76E8Xxniqd4C6T1sV5BuhKN1o?usp=sharing) | Flutter │ │
│ │ │(https://github.com/sayannath/American-Sign-Language-Detection) │ │
│Classificatio│EfficientNet-Lite0 (download │Icon Classifier Colab & Android (https://github.com/margaretmz/icon-classifier) | tutorial 1 │Community │
│n │(https://github.com/margaretmz/icon-classifier│(https://medium.com/swlh/icon-classifier-with-tflite-model-maker-9263c0021f72) | tutorial 2 │ │
│ │/blob/master/ml-code/icons-50.tflite)) │(https://medium.com/@margaretmz/icon-classifier-android-app-1fc0b727f761) │ │
Detection
│ Task │ Model │ App | Reference │ Source │
├─────────────────┼───────────────────────────────────────────────────────────────────────────────────┼──────────────────────────────────────────────────────────────────┼────────────────────┤
│Object detection │Quantized COCO SSD MobileNet v1 (download │Android │tensorflow.org │
│ │(https://storage.googleapis.com/download.tensorflow.org/models/tflite/coco_ssd_mobi│(https://github.com/tensorflow/examples/tree/master/lite/examples/│ │
│ │lenet_v1_1.0_quant_2018_06_29.zip)) │object_detection/android) | iOS │ │
│ │ │(https://github.com/tensorflow/examples/tree/master/lite/examples/│ │
│ │ │object_detection/ios) | Overview │ │
│ │ │(https://www.tensorflow.org/lite/models/object_detection/overview#│ │
│ │ │starter_model) │ │
│Object detection │YOLO │Flutter │Community │
│ │ │(https://blog.francium.tech/real-time-object-detection-on-mobile-w│ │
│ │ │ith-flutter-tensorflow-lite-and-yolo-android-part-a0042c9b62c6) | │ │
│ │ │Paper (https://arxiv.org/abs/1506.02640) │ │
│Object detection │YOLOv5 (https://tfhub.dev/neso613/lite-model/yolo-v5-tflite/tflite_model/1) │Yolov5 Inference │Community │
│ │ │ (https://github.com/neso613/yolo-v5-tflite-model) │ │
│Object detection │MobileNetV2 SSD (download │Reference │MediaPipe │
│ │(https://github.com/google/mediapipe/tree/master/mediapipe/models/ssdlite_object_de│(https://github.com/google/mediapipe/blob/master/mediapipe/models/│ │
│ │tection.tflite)) │object_detection_saved_model/README.md) │ │
│Object detection │MobileDet (Paper (https://arxiv.org/abs/2004.14525)) │Blog post (includes the TFLite conversion process) │MobileDet is from │
│ │ │ (https://sayak.dev/mobiledet-optimization/) │University of │
│ │ │ │Wisconsin-Madison │
│ │ │ │and Google and the │
│ │ │ │blog post is from │
│ │ │ │the Community │
│License Plate │SSD MobileNet (download) │Flutter (https://github.com/ariG23498/Flutter-License) │Community │
│detection │(https://github.com/ariG23498/Flutter-License/blob/master/assets/detect.tflite) │ │ │
│Face detection │BlazeFace (download │Paper │MediaPipe │
│ │(https://github.com/google/mediapipe/tree/master/mediapipe/models/face_detection_fr│ (https://sites.google.com/corp/view/perception-cv4arvr/blazeface)│ │
│ │ont.tflite)) │ │ │
│Face │FaceNet (https://arxiv.org/pdf/1503.03832.pdf) │Flutter (https://github.com/sayannath/Face-Authentication-App) │Community │
│Authentication │ │ │ │
│Hand detection & │Palm detection & hand landmarks (download │Blog post (https://mediapipe.page.link/handgoogleaiblog) | Model │MediaPipe & │
│tracking │(https://github.com/google/mediapipe/tree/master/mediapipe/models#hand-detection-an│card (https://mediapipe.page.link/handmc) | Android │Community │
│ │d-tracking)) │(https://github.com/supremetech/mediapipe-demo-hand-detection) │ │
Segmentation
│ Task │ Model │ App | Reference │ Source │
├────────────┼───────────────────────────────────────────────────────────────────────────────────────────┼────────────────────────────────────────────────────────────────────────┼───────────┤
│Segmentation│DeepLab V3 (download │Android & iOS │tf.org & │
│ │(https://storage.googleapis.com/download.tensorflow.org/models/tflite/gpu/deeplabv3_257_mv_│(https://github.com/tensorflow/examples/tree/master/lite/examples/image_│Community │
│ │gpu.tflite)) │segmentation/) | Overview │ │
│ │ │(https://www.tensorflow.org/lite/models/segmentation/overview) | Flutter│ │
│ │ │Image │ │
│ │ │(https://github.com/kshitizrimal/Flutter-TFLite-Image-Segmentation) | │ │
│ │ │Realtime (https://github.com/kshitizrimal/tflite-realtime-flutter) | │ │
│ │ │Paper (https://arxiv.org/abs/1706.05587) │ │
│Segmentation│Different variants of DeepLab V3 models │Models on TF Hub │Community │
│ │(https://github.com/tensorflow/models/blob/master/research/deeplab/g3doc/model_zoo.md) │(https://tfhub.dev/s?module-type=image-segmentation&publisher=sayakpaul)│ │
│ │ │with Colab Notebooks │ │
│Segmentation│DeepLab V3 model │Android (https://github.com/farmaker47/Update_image_segmentation) | │Community │
│ │ (https://tfhub.dev/tensorflow/lite-model/deeplabv3/1/metadata/2?lite-format=tflite) │Tutorial │ │
│ │ │(https://farmaker47.medium.com/use-camerax-with-image-segmentation-andro│ │
│ │ │id-project-d8656f35cea3) │ │
│Hair │Download │Paper │MediaPipe │
│Segmentation│(https://github.com/google/mediapipe/tree/master/mediapipe/models/hair_segmentation.tflite)│(https://sites.google.com/corp/view/perception-cv4arvr/hair-segmentation│ │
│ │ │) │ │
Style Transfer
│ Task │ Model │ App | Reference │ Source │
├─────────────────┼─────────────────────────────────────────────────────────────────────────────────────┼─────────────────────────────────────────────────────────────────────────┼───────────┤
│Style transfer │Arbitrary image stylization │Overview (https://www.tensorflow.org/lite/models/style_transfer/overview)│tf.org & │
│ │(https://github.com/tensorflow/magenta/tree/master/magenta/models/arbitrary_image_sty│| Android │Community │
│ │lization) │(https://github.com/tensorflow/examples/tree/master/lite/examples/style_t│ │
│ │ │ransfer/android) | Flutter │ │
│ │ │(https://github.com/PuzzleLeaf/flutter_tflite_style_transfer) │ │
│Style transfer │Better-quality style transfer models in .tflite │Models on TF Hub │Community │
│ │ │(https://tfhub.dev/sayakpaul/lite-model/arbitrary-image-stylization-incep│ │
│ │ │tionv3/dr/predict/1) with Colab Notebooks │ │
│Video Style │Download: Dynamic range models │Android (https://github.com/farmaker47/video_style_transfer) | Tutorial │Community │
│Transfer │(https://tfhub.dev/sayakpaul/lite-model/arbitrary-image-stylization-inceptionv3-dynam│(https://medium.com/@farmaker47/android-implementation-of-video-style-tra│ │
│ │ic-shapes/dr/transfer/1)) │nsfer-with-tensorflow-lite-models-9338a6d2a3ea) │ │
│Segmentation & │DeepLabV3 & Style Transfer models │Project repo (https://github.com/margaretmz/segmentation-style-transfer)|│Community │
│Style transfer │(https://github.com/margaretmz/segmentation-style-transfer/tree/master/ml) │Android │ │
│ │ │(https://github.com/margaretmz/segmentation-style-transfer/tree/master/an│ │
│ │ │droid) | Tutorial │ │
│ │ │(https://medium.com/google-developer-experts/image-background-stylizer-pa│ │
│ │ │rt-1-project-intro-d68c4547e7e3) │ │
Generative
│ Task │ Model │ App | Reference │ Source │
├──────────┼───────────────────────────────────────────────────────────────────────────┼────────────────────────────────────────────────────────────────────────────────────────────┼─────────┤
│GANs │U-GAT-IT (https://github.com/taki0112/UGATIT) (Selfie2Anime) │Project repo (https://github.com/margaretmz/selfie2anime-with-tflite) | Android │Community│
│ │ │(https://github.com/margaretmz/selfie2anime-with-tflite/tree/master/android) | Tutorial │ │
│ │ │(https://medium.com/google-developer-experts/selfie2anime-with-tflite-part-1-overview-f97500│ │
│ │ │800ffe) │ │
│GANs │White-box CartoonGAN │Project repo (https://github.com/margaretmz/Cartoonizer-with-TFLite) | Android │Community│
│ │(https://github.com/SystemErrorWang/White-box-Cartoonization) (download │(https://github.com/margaretmz/Cartoonizer-with-TFLite/tree/master/android) | Tutorial │ │
│ │(https://tfhub.dev/sayakpaul/lite-model/cartoongan/dr/1)) │(https://blog.tensorflow.org/2020/09/how-to-create-cartoonizer-with-tf-lite.html) │ │
│GANs - │Boundless on TF Hub │Colab Notebook │Community│
│Image │(https://tfhub.dev/sayakpaul/lite-model/boundless-quarter/dr/1) │(https://colab.research.google.com/github/sayakpaul/Adventures-in-TensorFlow-Lite/blob/maste│ │
│Extrapolat│ │r/Boundless_TFLite.ipynb) | Original Paper (https://arxiv.org/pdf/2003.06792v2.pdf) │ │
│ion │ │ │ │
Post estimation
│ Task │ Model │ App | Reference │ Source │
├─────────────────────┼────────────────────────────────────────────────────────────────────────────────────────────────────────┼───────────────────────────────────────────────────┼──────────┤
│Pose estimation │Posenet (download │Android │tensorflow│
│ │(https://storage.googleapis.com/download.tensorflow.org/models/tflite/posenet_mobilenet_v1_100_257x257_m│(https://github.com/tensorflow/examples/tree/master│.org │
│ │ulti_kpt_stripped.tflite)) │/lite/examples/posenet/android) | iOS │ │
│ │ │(https://github.com/tensorflow/examples/tree/master│ │
│ │ │/lite/examples/posenet/ios) | Overview │ │
│ │ │(https://www.tensorflow.org/lite/models/pose_estima│ │
│ │ │tion/overview) │ │
│Pose Classification │MoveNet Lightning (download │Project Repository │Community │
│based Video Game │(https://github.com/NSTiwari/Video-Game-Control-using-Pose-Classification-and-TensorFlow-Lite/blob/main/│(https://github.com/NSTiwari/Video-Game-Control-usi│ │
│Control │movenet_lightning.tflite)) │ng-Pose-Classification-and-TensorFlow-Lite) │ │
Other
│ Task │ Model │ App | Reference │ Source │ │
├───────────────────────────┼────────────────────────────────────────────────────────┼────────────────────────────────────────────────────────────────────────────────────┼─────────┼─────────┤
│Low-light image enhancement│Models on TF Hub │Project repo (https://github.com/sayakpaul/MIRNet-TFLite) | Original Paper │ │Community│
│ │ (https://tfhub.dev/sayakpaul/mirnet-fixed/1) │(https://arxiv.org/pdf/2003.06792v2.pdf) | Flutter │ │ │
│ │ │(https://github.com/sayannath/MIRNet-Flutter) │ │ │
│OCR │Models on TF Hub │Project Repository (https://github.com/tulasiram58827/ocr_tflite) │Community│ │
│ │(https://tfhub.dev/tulasiram58827/lite-model/keras-ocr/d│ │ │ │
│ │r/2) │ │ │ │
Text
│ Task │ Model │ Sample apps │ Source │
├──────────────┼───────────────────────────────────────────────┼──────────────────────────────────────────────────────────────────────────────────────────────────────────────────┼───────────┤
│Question & │DistilBERT │Android (https://github.com/huggingface/tflite-android-transformers/blob/master/bert) │Hugging │
│Answer │ │ │Face │
│Text │GPT-2 / DistilGPT2 │Android (https://github.com/huggingface/tflite-android-transformers/blob/master/gpt2) │Hugging │
│Generation │ │ │Face │
│Text │Download │Android (https://github.com/tensorflow/examples/tree/master/lite/examples/text_classification/android) |iOS │tf.org & │
│Classification│(https://storage.googleapis.com/download.tensor│(https://github.com/khurram18/TextClassafication) | Flutter │Community │
│ │flow.org/models/tflite/text_classification/text│(https://github.com/am15h/tflite_flutter_plugin/tree/master/example) │ │
│ │_classification.tflite) │ │ │
│Text Detection│CRAFT Text Detector (Paper │Download (https://github.com/tulasiram58827/craft_tflite/blob/main/models/craft_float_800.tflite?raw=true) | │Community │
│ │(https://arxiv.org/pdf/1904.01941)) │Project Repository (https://github.com/tulasiram58827/craft_tflite/) | Blog1-Conversion to TFLite │ │
│ │ │(https://tulasi.dev/craft-in-tflite) | Blog2-EAST vs CRAFT (https://sayak.dev/optimizing-text-detectors/) | Models│ │
│ │ │on TF Hub (https://tfhub.dev/tulasiram58827/lite-model/craft-text-detector/dr/1) | Android (Coming Soon) │ │
│Text Detection│EAST Text Detector (Paper │Models on TF Hub (https://tfhub.dev/sayakpaul/lite-model/east-text-detector/dr/1) | Conversion and Inference │Community │
│ │(https://arxiv.org/abs/1704.03155)) │Notebook │ │
│ │ │(https://colab.research.google.com/github/sayakpaul/Adventures-in-TensorFlow-Lite/blob/master/EAST_TFLite.ipynb) │ │
Speech
│ Task │ Model │ App | Reference │ Source │
├─────────────────────┼───────────────────────────────────────────────────────────────┼──────────────────────────────────────────────────────────────────────────────────────────┼────────────┤
│Speech Recognition │DeepSpeech │Reference (https://github.com/mozilla/DeepSpeech/tree/master/native_client/java) │Mozilla │
│Speech Recognition │CONFORMER │Inference (https://github.com/neso613/ASR_TFLite) Android │Community │
│ │ │(https://github.com/windmaple/tflite-asr) │ │
│Speech Synthesis │Tacotron-2, FastSpeech2, MB-Melgan │Android (https://github.com/TensorSpeech/TensorflowTTS/tree/master/examples/android) │TensorSpeech│
│Speech Synthesis(TTS)│Tacotron2, FastSpeech2, MelGAN, MB-MelGAN, HiFi-GAN, Parallel │Inference Notebook │Community │
│ │WaveGAN │(https://github.com/tulasiram58827/TTS_TFLite/blob/main/End_to_End_TTS.ipynb) | │ │
│ │ │Project Repository (https://github.com/tulasiram58827/TTS_TFLite/) │ │
Recommendation
│ Task │ Model │ App | Reference │ Source │
├──────────────┼───────────────────────────────────────────────────┼─────────────────────────────────────────────────────────────────────────────────────────────────────────────┼────────────┤
│On-device │Dual-Encoder │Android (https://github.com/tensorflow/examples/tree/master/lite/examples/recommendation/android) | iOS │tf.org & │
│Recommendation│(https://github.com/tensorflow/examples/tree/master│(https://github.com/zhuzilin/on-device_recommendation_tflite) | Reference │Community │
│ │/lite/examples/recommendation/ml) │(https://blog.tensorflow.org/2020/09/introduction-to-tflite-on-device-recommendation.html) │ │
Game
│ Task │ Model │ App | Reference │ Source │
├──────────┼──────────────────────┼─────────────────────────────────────────────────────────────────────────────────────────────────────┼─────────┤
│Game agent│Reinforcement learning│Flutter (https://github.com/windmaple/planestrike-flutter) | Tutorial (https://windmaple.medium.com/)│Community│
Model zoo
TensorFlow Lite models
These are the TensorFlow Lite models that could be implemented in apps and things:
⟡ MobileNet (https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet/README.md) - Pretrained MobileNet v2 and v3 models.
⟡ TensorFlow Lite models
⟡ TensorFlow Lite models (https://www.tensorflow.org/lite/models) - With official Android and iOS examples.
⟡ Pretrained models (https://www.tensorflow.org/lite/guide/hosted_models) - Quantized and floating point variants.
⟡ TensorFlow Hub (https://tfhub.dev/) - Set "Model format = TFLite" to find TensorFlow Lite models.
TensorFlow models
These are TensorFlow models that could be converted to .tflite and then implemented in apps and things:
⟡ TensorFlow models (https://github.com/tensorflow/models/tree/master/official) - Official TensorFlow models.
⟡ Tensorflow detection model zoo
(https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/tf2_detection_zoo.md) - Pre-trained on COCO, KITTI, AVA v2.1, iNaturalist Species datasets.
Ideas and Inspiration
⟡ E2E TFLite Tutorials (https://github.com/ml-gde/e2e-tflite-tutorials) - Checkout this repo for sample app ideas and seeking help for your tutorial projects. Once a project gets completed,
the links of the TensorFlow Lite model(s), sample code and tutorial will be added to this awesome list.
ML Kit examples
ML Kit (https://developers.google.com/ml-kit) is a mobile SDK that brings Google's ML expertise to mobile developers.
⟡ 2019-10-01 ML Kit Translate demo (https://codelabs.developers.google.com/codelabs/mlkit-android-translate/#0) - A tutorial with material design Android
(https://github.com/googlecodelabs/mlkit-android/tree/master/translate) (Kotlin) sample - recognize, identify Language and translate text from live camera with ML Kit for Firebase.
⟡ 2019-03-13 Computer Vision with ML Kit - Flutter In Focus (https://youtu.be/ymyYUCrJnxU).
⟡ 2019-02-09 Flutter + MLKit: Business Card Mail Extractor (https://medium.com/flutter-community/flutter-mlkit-8039ec66b6a) - A blog post with a Flutter
(https://github.com/DaemonLoki/Business-Card-Mail-Extractor) sample code.
⟡ 2019-02-08 From TensorFlow to ML Kit: Power your Android application with machine learning
(https://speakerdeck.com/jinqian/from-tensorflow-to-ml-kit-power-your-android-application-with-machine-learning) - A talk with Android (https://github.com/xebia-france/magritte) (Kotlin)
sample code.
⟡ 2018-08-07 Building a Custom Machine Learning Model on Android with TensorFlow Lite
(https://medium.com/over-engineering/building-a-custom-machine-learning-model-on-android-with-tensorflow-lite-26447e53abf2).
⟡ 2018-07-20 ML Kit and Face Detection in Flutter (https://flatteredwithflutter.com/ml-kit-and-face-detection-in-flutter/).
⟡ 2018-07-27 ML Kit on Android 4: Landmark Detection (https://medium.com/google-developer-experts/exploring-firebase-mlkit-on-android-landmark-detection-part-four-5e86b8deac3a).
⟡ 2018-07-28 ML Kit on Android 3: Barcode Scanning (https://medium.com/google-developer-experts/exploring-firebase-mlkit-on-android-barcode-scanning-part-three-cc6f5921a108).
⟡ 2018-05-31 ML Kit on Android 2: Face Detection (https://medium.com/google-developer-experts/exploring-firebase-mlkit-on-android-face-detection-part-two-de7e307c52e0).
⟡ 2018-05-22 ML Kit on Android 1: Intro (https://medium.com/google-developer-experts/exploring-firebase-mlkit-on-android-introducing-mlkit-part-one-98fcfedbeee0).
Plugins and SDKs
⟡ Edge Impulse (https://www.edgeimpulse.com/) - Created by @EdgeImpulse (https://twitter.com/EdgeImpulse) to help you to train TensorFlow Lite models for embedded devices in the cloud.
⟡ MediaPipe (https://github.com/google/mediapipe) - A cross platform (mobile, desktop and Edge TPUs) AI pipeline by Google AI. (PM Ming Yong (https://twitter.com/realmgyong)) | MediaPipe
examples (https://mediapipe.readthedocs.io/en/latest/examples.html).
⟡ Coral Edge TPU (https://coral.ai/) - Edge hardware by Google. Coral Edge TPU examples (https://coral.ai/examples/).
⟡ TensorFlow Lite Flutter Plugin (https://github.com/am15h/tflite_flutter_plugin/) - Provides a dart API similar to the TensorFlow Lite Java API for accessing TensorFlow Lite interpreter and
performing inference in flutter apps. tflite_flutter on pub.dev (https://pub.dev/packages/tflite_flutter).
Helpful links
⟡ Netron (https://github.com/lutzroeder/netron) - A tool for visualizing models.
⟡ AI benchmark (http://ai-benchmark.com/tests.html) - A website for benchmarking computer vision models on smartphones.
⟡ Performance measurement (https://www.tensorflow.org/lite/performance/measurement) - How to measure model performance on Android and iOS.
⟡ Material design guidelines for ML (https://material.io/collections/machine-learning/patterns-for-machine-learning-powered-features.html) - How to design machine learning powered features. A
good example: ML Kit Showcase App (https://github.com/firebase/mlkit-material-android).
⟡ The People + AI Guide book (https://pair.withgoogle.com/) - Learn how to design human-centered AI products.
⟡ Adventures in TensorFlow Lite
(https://github.com/sayakpaul/Adventures-in-TensorFlow-Lite) - A repository showing non-trivial conversion processes and general explorations in TensorFlow Lite.
⟡ TFProfiler (https://github.com/iglaweb/TFProfiler) - An Android-based app to profile TensorFlow Lite models and measure its performance on smartphone.
⟡ TensorFlow Lite for Microcontrollers (https://www.tensorflow.org/lite/microcontrollers)
⟡ TensorFlow Lite Examples - Android (https://github.com/dailystudio/tensorflow-lite-examples-android) - A repository refactors and rewrites all the TensorFlow Lite Android examples which are
included in the TensorFlow official website.
⟡ Tensorflow-lite-kotlin-samples (https://github.com/SunitRoy2703/Tensorflow-lite-kotlin-samples) - A collection of Tensorflow Lite Android example Apps in Kotlin, to show different kinds of
kotlin implementation of the example apps (https://www.tensorflow.org/lite/examples)
Learning resources
Interested but not sure how to get started? Here are some learning resources that will help you whether you are a beginner or a practitioner in the field for a while.
Blog posts
⟡ 2021-11-09 On-device training in TensorFlow Lite (https://blog.tensorflow.org/2021/11/on-device-training-in-tensorflow-lite.html)
⟡ 2021-09-27 Optical character recognition with TensorFlow Lite: A new example app (https://blog.tensorflow.org/2021/09/blog.tensorflow.org202109optical-character-recognition.html)
⟡ 2021-06-16 https://blog.tensorflow.org/2021/06/easier-object-detection-on-mobile-with-tf-lite.html (https://blog.tensorflow.org/2021/11/on-device-training-in-tensorflow-lite.html)
⟡ 2020-12-29 YOLOv3 to TensorFlow Lite Conversion (https://medium.com/analytics-vidhya/yolov3-to-tensorflow-lite-conversion-4602cec5c239) - By Nitin Tiwari.
⟡ 2020-04-20 What is new in TensorFlow Lite (https://blog.tensorflow.org/2020/04/whats-new-in-tensorflow-lite-from-devsummit-2020.html) - By Khanh LeViet.
⟡ 2020-04-17 Optimizing style transfer to run on mobile with TFLite (https://blog.tensorflow.org/2020/04/optimizing-style-transfer-to-run-on-mobile-with-tflite.html) - By Khanh LeViet and
Luiz Gustavo Martins.
⟡ 2020-04-14 How TensorFlow Lite helps you from prototype to product (https://blog.tensorflow.org/2020/04/how-tensorflow-lite-helps-you-from-prototype-to-product.html) - By Khanh LeViet.
⟡ 2019-11-08 Getting Started with ML on MCUs with TensorFlow (https://blog.particle.io/2019/11/08/particle-machine-learning-101/) - By Brandon Satrom.
⟡ 2019-08-05 TensorFlow Model Optimization Toolkit — float16 quantization halves model size (https://blog.tensorflow.org/2019/08/tensorflow-model-optimization-toolkit_5.html) - By the
TensorFlow team.
⟡ 2018-07-13 Training and serving a real-time mobile object detector in 30 minutes with Cloud TPUs
(https://blog.tensorflow.org/2018/07/training-and-serving-realtime-mobile-object-detector-cloud-tpus.html) - By Sara Robinson, Aakanksha Chowdhery, and Jonathan Huang.
⟡ 2018-06-11 - Why the Future of Machine Learning is Tiny (https://petewarden.com/2018/06/11/why-the-future-of-machine-learning-is-tiny/) - By Pete Warden.
⟡ 2018-03-30 - Using TensorFlow Lite on Android (https://blog.tensorflow.org/2018/03/using-tensorflow-lite-on-android.html)) - By Laurence Moroney.
Books
⟡ 2021-12-01 AI and Machine Learning On-Device Development (https://learning.oreilly.com/library/view/ai-and-machine/9781098101732/) (early access) - By Laurence Moroney (@lmoroney
(https://twitter.com/lmoroney)).
⟡ 2020-10-01 AI and Machine Learning for Coders (https://learning.oreilly.com/library/view/ai-and-machine/9781492078180/) - By Laurence Moroney (@lmoroney (https://twitter.com/lmoroney)).
⟡ 2020-04-06 Mobile Deep Learning with TensorFlow Lite, ML Kit and Flutter (https://www.packtpub.com/product/mobile-deep-learning-with-tensorflow-lite-ml-kit-and-flutter/9781789611212): Build
scalable real-world projects to implement end-to-end neural networks on Android and iOS (GitHub (https://github.com/PacktPublishing/Mobile-Deep-Learning-Projects)) - By Anubhav Singh (
@xprilion (https://github.com/xprilion)) and Rimjhim Bhadani (@Rimjhim28 (https://github.com/Rimjhim28)).
⟡ 2020-03-01 Raspberry Pi for Computer Vision (Complete Bundle (https://www.pyimagesearch.com/raspberry-pi-for-computer-vision) | TOC
(https://www.pyimagesearch.com/2019/04/05/table-of-contents-raspberry-pi-for-computer-vision/)) - By the PyImageSearch Team: Adrian Rosebrock (@PyImageSearch
(https://twitter.com/PyImageSearch)), David Hoffman, Asbhishek Thanki, Sayak Paul (@RisingSayak (https://twitter.com/RisingSayak)), and David Mcduffee.
⟡ 2019-12-01 TinyML (http://shop.oreilly.com/product/0636920254508.do) - By Pete Warden (@petewarden (https://twitter.com/petewarden)) and Daniel Situnayake (@dansitu
(https://twitter.com/dansitu)).
⟡ 2019-10-01 Practical Deep Learning for Cloud, Mobile, and Edge (https://www.practicaldeeplearning.ai/) - By Anirudh Koul (@AnirudhKoul (https://twitter.com/AnirudhKoul)), Siddha Ganju (
@SiddhaGanju (https://twitter.com/SiddhaGanju)), and Meher Kasam (@MeherKasam (https://twitter.com/MeherKasam)).
Videos
⟡ 2021-10-06 Contributing to TensorFlow Lite with Sunit Roy (https://youtu.be/sZayUoWW6nE) (Hacktoberfest 2021)
⟡ 2020-07-25 Android ML by Hoi Lam (https://youtu.be/m_bEh8YifnQ) (GDG Kolkata meetup).
⟡ 2020-04-01 Easy on-device ML from prototype to production (https://youtu.be/ALxWJoh_BHw) (TF Dev Summit 2020).
⟡ 2020-03-11 TensorFlow Lite: ML for mobile and IoT devices (https://youtu.be/27Zx-4GOQA8) (TF Dev Summit 2020).
⟡ 2019-10-31 Keynote - TensorFlow Lite: ML for mobile and IoT devices (https://youtu.be/zjDGAiLqGk8).
⟡ 2019-10-31 TensorFlow Lite: Solution for running ML on-device (https://youtu.be/0SpZy7iouFU).
⟡ 2019-10-31 TensorFlow model optimization: Quantization and pruning (https://youtu.be/3JWRVx1OKQQ).
⟡ 2019-10-29 Inside TensorFlow: TensorFlow Lite (https://youtu.be/gHN0jDbJz8E).
⟡ 2018-04-18 TensorFlow Lite for Android (Coding TensorFlow) (https://youtu.be/JnhW5tQ_7Vo).
Podcasts
⟡ 2020-08-08 Talking Machine Learning with Hoi Lam (https://anchor.fm/talkingwithapples/episodes/Talking-Machine-Learning-with-Hoi-Lam-eiaj7v).
MOOCs
⟡ Introduction to TensorFlow Lite (https://www.udacity.com/course/intro-to-tensorflow-lite--ud190) - Udacity course by Daniel Situnayake (@dansitu), Paige Bailey (@DynamicWebPaige
(https://twitter.com/DynamicWebPaige)), and Juan Delgado.
⟡ Device-based Models with TensorFlow Lite (https://www.coursera.org/learn/device-based-models-tensorflow) - Coursera course by Laurence Moroney (@lmoroney (https://twitter.com/lmoroney)).
⟡ The Future of ML is Tiny and Bright (https://www.edx.org/professional-certificate/harvardx-tiny-machine-learning) - A series of edX courses created by Harvard in collaboration with Google.
Instructors - Vijay Janapa Reddi, Laurence Moroney, and Pete Warden.