Internal change

PiperOrigin-RevId: 476240601

official/projects/waste_identification_ml/README.md

 # CircularNet
 
-Instance segmentation models for identification of recyclables on conveyor belts
+Instance segmentation models for identification of recyclables on conveyor
+belts.
 
-## Code Structure
-This is an implementation of Mask RCNN based on Python 3 and Tensorflow 2.x. The
-model generates bounding boxes and segmentation masks for each instance of an
-object in the image. The repository includes :
+Note that these are demo models, reach out to
+waste-innovation-external@google.com for updated versions of the models.
 
-* Source code for training a Mask RCNN model.
-* Inference code
-* Pre-trained weights for inferencing
-* Docker to deploy the model in any operating system and run.
-* Jupyter notebook to visualize the detection pipeline at every step.
-* Evaluation metric of the validation dataset.
-* Example of training on your own custom dataset.
+## Overview
 
-The code is designed in such a way so that it can be extended. If you use it in
-your research or industrial solutions, then please consider citing this
-repository.
+Mask RCNN is a state-of-art deep learning model for instance image segmentation,
+where the goal is to assign instance level labels ( e.g. person1, person2, cat)
+to every pixel in an input image. Mask RCNN algorithm is available in the
+TensorFlow Model Garden which is a repository with a number of different
+implementations of state-of-the-art models and modeling solutions for TensorFlow
+users.
 
-## Pre-requisites
+## Model Categories
 
-## Prepare dataset
+-   Material model - Detects the high level category (e.g. plastic, paper, etc)
+    of an object according to its material type.
+-   Material Form model - Detects the category of the of an object according to
+    its physical product formation (e.g. cup, plate, pen, etc).
+-   Plastic model - Detects the category of a object according to its plastic
+    types (e.g. HDPE, LDPE, etc)
 
-## Setup virtual systems for training
+> The goal to develop these models is to bring transparency & traceability in
+the world of  waste recycling.
 
-### ***Start a TPU v3-32 instance***
+## Model paths in GCP buckets
 
--   [x] Set up a Google cloud account on GCP
--   [x] Go to the cloud console and create a new project.
--   [x] While setting up your project, you will be asked to set up a billing
-    account. You will only be charged after you start using it.
--   [x] Create a cloud TPU project
--   [x] Link for the above 4 steps can be
-    [found here](https://cloud.google.com/tpu/docs/setup-gcp-account)
--   [x] Once the project is created, select the project from the cloud console.
--   [x] On the top right, click cloud shell to open the terminal. See
-    [TPU Quickstart](https://cloud.google.com/tpu/docs/quick-starts) for
-    instructions.
-    An example command would look like:
-    ```bash
-    ctpu up --name
-    <tpu-name> --zone <zone> --tpu-size=v3-32 --tf-version nightly --project
-    <project ID>
-    ```
-    **Example** -
-
--   This model requires TF version >= 2.5. Currently, that is only available via
-    a nightly build on Cloud.
-
--   You can check TPU types with their cores and memory
-    [here](https://cloud.google.com/tpu/docs/types-zones#tpu-vm) and select
-    accordingly.
-
--   CAREFULLY choose a TPU type which can be turned ON and OFF after usage.
-    The preferred one is below - `bash ctpu up --name waste-identification --zone
-    us-central1-a --tpu-size=v3-8 --tf-version nightly --project
-    waste-identification-ml` After the execution of the above command, you will
-    see 2 virtual devices with name "waste-identification" each in TPU and
-    COMPUTE ENGINE section.
-
-### ***Get into the virtual machine***
-
-The virtual machine, which is a TPU host, can be seen in the COMPUTE ENGINE
-section of GCP. We will use this virtual machine to start the training process.
-This machine will use another virtual instance of TPU that is found in the TPU
-section of the GCP. To get inside the TPU host virtual machine :
-
--   Go to the COMPUTE ENGINE section in the GCP
--   Find your instance there
--   Under the "Connect" tab of your instance, you will see "SSH",
--   Click on SSH and it will open another window which will take you inside the
-    virtual machine.
--   Use the following commands inside the virtual machine window :
-
-```bash
-$ git clone https://github.com/tensorflow/models.git
-$ cd models
-$ pip3 install -r official/requirements.txt
-``
+| Model categories | Model backbone | Model type | GCP bucket path |
+| ------ | ------ | ----- | ------ |
+| Material Model | Resnet | saved model & TFLite | [click here](https://storage.googleapis.com/tf_model_garden/vision/waste_identification_ml/material_model.zip) |
+| Material Form model | Resnet | saved model & TFLite | [click here](https://storage.googleapis.com/tf_model_garden/vision/waste_identification_ml/material_form_model.zip) |
+|Plastic model | Resnet| saved model & TFLite | [click here](https://storage.googleapis.com/tf_model_garden/vision/waste_identification_ml/plastic_types_model.zip) |

official/projects/waste_identification_ml/pre_trained_models.md

-# Instance Segmentation Model Weights
-
-Light weight deep learning models for instance image segmentation.
-## Overview
-
-Mask RCNN is a state-of-art deep learning model for instance image segmentation, where the goal is to assign instance level labels ( e.g. person1, person2, cat) to every pixel in an input image. Mask RCNN algorithm is available in the TensorFlow Model Garden which is a repository with a number of different implementations of state-of-the-art models and modeling solutions for TensorFlow users.
-
-## Model Categories
-
-- Material model - Detects the high level category (e.g. plastic, paper, etc) of an object according to its material type.
-- Material Form model - Detects the category of the of an object according to its physical product formation (e.g. cup, plate, pen, etc).
-- Plastic model - Detects the category of a object according to its plastic types (e.g. HDPE, LDPE, etc)
-
-> The goal to develop these models is to bring transparency & traceability in the world of  waste recycling.
-
-## Model paths in GCP buckets
-
-| Model categories | Model backbone | Model type | GCP bucket path |
-| ------ | ------ | ----- | ------ |
-| Material Model | Resnet | saved model & TFLite | [click here](https://storage.googleapis.com/tf_model_garden/vision/waste_identification_ml/material_model.zip) |
-| Material Form model | Resnet | saved model & TFLite | [click here](https://storage.googleapis.com/tf_model_garden/vision/waste_identification_ml/material_form_model.zip) |
-|Plastic model | Resnet| saved model & TFLite | [click here](https://storage.googleapis.com/tf_model_garden/vision/waste_identification_ml/plastic_types_model.zip) |