ibm / max-scene-classifier

Image classifier for physical places/locations, based on the Places365-CNN Model

IBM Code Model Asset Exchange: Scene Classifier

This repository contains code to instantiate and deploy an image classification model. This model recognizes the 365 different classes of scene/location in the Places365-Standard subset of the Places2 Dataset. The model is based on the Places365-CNN Model and consists of a pre-trained deep convolutional net using the ResNet architecture, trained on the ImageNet-2012 data set. The pre-trained model is then fine-tuned on the Places365-Standard dataset. The input to the model is a 224x224 image, and the output is a list of estimated class probabilities.

The specific model variant used in this repository is the PyTorch Places365 ResNet18 Model. The model files are hosted on IBM Cloud Object Storage. The code in this repository deploys the model as a web service in a Docker container. This repository was developed as part of the IBM Code Model Asset Exchange and the public API is powered by IBM Cloud.

Model Metadata

Domain	Application	Industry	Framework	Training Data	Input Data Format
Vision	Image Classification	General	Pytorch	Places365	Image (RGB/HWC)

References

• B. Zhou, A. Lapedriza, A. Khosla, A. Oliva, and A. Torralba, "Places: A 10 million Image Database for Scene Recognition", IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017.
• B. Zhou, A. Lapedriza, J. Xiao, A. Torralba and A. Oliva, "Learning Deep Features for Scene Recognition using Places Database", Advances in Neural Information Processing Systems 27, 2014.
• K. He, X. Zhang, S. Ren and J. Sun, "Deep Residual Learning for Image Recognition", CoRR (abs/1512.03385), 2015.
• Places2 Project Page
• Places365-CNN GitHub Page

Licenses

Component	License	Link
This repository	Apache 2.0	LICENSE
Model Weights	CC BY License	Places365-CNN
Model Code (3rd party)	MIT	Places365-CNN
Test assets	Various	Asset README

Pre-requisites:

• docker: The Docker command-line interface. Follow the installation instructions for your system.
• The minimum recommended resources for this model is 2GB Memory and 2 CPUs.

Deployment options

• Deploy from Quay
• Deploy on Red Hat OpenShift
• Deploy on Kubernetes
• Run Locally

Deploy from Quay

To run the docker image, which automatically starts the model serving API, run:

$ docker run -it -p 5000:5000 quay.io/codait/max-scene-classifier

This will pull a pre-built image from the Quay.io container registry (or use an existing image if already cached locally) and run it. If you'd rather checkout and build the model locally you can follow the run locally steps below.

Deploy on Red Hat OpenShift

You can deploy the model-serving microservice on Red Hat OpenShift by following the instructions for the OpenShift web console or the OpenShift Container Platform CLI in this tutorial, specifying quay.io/codait/max-scene-classifier as the image name.

Deploy on Kubernetes

You can also deploy the model on Kubernetes using the latest docker image on Quay.

On your Kubernetes cluster, run the following commands:

$ kubectl apply -f https://raw.githubusercontent.com/IBM/MAX-Scene-Classifier/master/max-scene-classifier.yaml

The model will be available internally at port 5000, but can also be accessed externally through the NodePort.

A more elaborate tutorial on how to deploy this MAX model to production on IBM Cloud can be found here.

Run Locally

1. Build the Model
2. Deploy the Model
3. Use the Model
4. Development
5. Cleanup

1. Build the Model

Clone this repository locally. In a terminal, run the following command:

$ git clone https://github.com/IBM/MAX-Scene-Classifier.git

Change directory into the repository base folder:

$ cd MAX-Scene-Classifier

To build the docker image locally, run:

$ docker build -t max-scene-classifier .

All required model assets will be downloaded during the build process. Note that currently this docker image is CPU only (we will add support for GPU images later).

2. Deploy the Model

To run the docker image, which automatically starts the model serving API, run:

$ docker run -it -p 5000:5000 max-scene-classifier

3. Use the Model

The API server automatically generates an interactive Swagger documentation page. Go to http://localhost:5000 to load it. From there you can explore the API and also create test requests.

Use the model/predict endpoint to load a test image (you can use one of the test images from the samples folder) and get predicted labels for the image from the API.

You can also test it on the command line, for example:

$ curl -F "image=@samples/aquarium.jpg" -XPOST http://localhost:5000/model/predict

You should see a JSON response like that below:

{
  "status": "ok",
  "predictions": [
    {
      "label_id": "9",
      "label": "aquarium",
      "probability": 0.97350615262985
    },
    {
      "label_id": "342",
      "label": "underwater\/ocean_deep",
      "probability": 0.0062678409740329
    },
    {
      "label_id": "297",
      "label": "science_museum",
      "probability": 0.005441018845886
    },
    {
      "label_id": "239",
      "label": "natural_history_museum",
      "probability": 0.00413528829813
    },
    {
      "label_id": "167",
      "label": "grotto",
      "probability": 0.0024146677460521
    }
  ]
}

4. Development

To run the Flask API app in debug mode, edit config.py to set DEBUG = True under the application settings. You will then need to rebuild the docker image (see step 1).

5. Cleanup

To stop the Docker container, type CTRL + C in your terminal.

Resources and Contributions

If you are interested in contributing to the Model Asset Exchange project or have any queries, please follow the instructions here.

max-scene-classifier's Issues

Handle images with alpha channel data

/model/predict throws HTTP code 500 if an image is provided as input containing alpha channel data.

/model/predict: Handle invalid input gracefully

Currently if an invalid input is specified (e.g. invalid file format) the /model/predict API endpoint returns HTTP code 500 instead of 400.

pip install from requirements.txt

Example PR:
IBM/MAX-Review-Text-Generator#7

top k logic issue

I think it should be x.sort(1, True), and the results should be the same with torch.topk API.

MAX-Scene-Classifier/core/model.py

Line 93 in 0927fc5

probs, idx = x.sort(0, True)

Sample codes:

>>> import torch
>>> probs = torch.tensor([[0.0297, 0.3163, 0.4335, 0.2205],
                [0.3554, 0.2583, 0.3467, 0.0396],
                [0.3997, 0.4566, 0.0879, 0.0557]])

>>> probs.sort(1,True)

orch.return_types.sort(
values=tensor([[0.4335, 0.3163, 0.2205, 0.0297],
        [0.3554, 0.3467, 0.2583, 0.0396],
        [0.4566, 0.3997, 0.0879, 0.0557]]),
indices=tensor([[2, 1, 3, 0],
        [0, 2, 1, 3],
        [1, 0, 2, 3]]))

>>> torch.topk(probs, 4)

torch.return_types.topk(
values=tensor([[0.4335, 0.3163, 0.2205, 0.0297],
        [0.3554, 0.3467, 0.2583, 0.0396],
        [0.4566, 0.3997, 0.0879, 0.0557]]),
indices=tensor([[2, 1, 3, 0],
        [0, 2, 1, 3],
        [1, 0, 2, 3]]))

Move to MAX-Framework

Update model to use https://github.com/IBM/MAX-Framework

Standardize model file packaging

Refer to https://github.ibm.com/CODAIT/Production-Readiness/issues/4

This model does not package the model files into tar.gz like TensorFlow models.

inconsistent link to COS directory for model files

the README contains this link to the model files:

• http://max-assets.s3-api.us-geo.objectstorage.softlayer.net/pytorch/places365/whole_resnet18_places365_python36.pth

this works. however, it is different than the link in the Dockerfile:

• http://max-assets.s3-api.us-geo.objectstorage.softlayer.net/scene-classifier/1.0/assets.tar.gz

i would expect both of these to point to the same set of model files.
in addition, both of the links should be pointing to the new US cross region:

s3.us.cloud-object-storage.appdomain.cloud

because the old one has been deprecated:

s3-api.us-geo.objectstorage.softlayer.net

Add request param descriptions and update Swagger models help text

Example in PR:
IBM/MAX-Object-Detector#20

Fix the spelling of aquarium

• rename /assets/acquarium.jpg
• update /assets/README.md, which references the image
• update /README.md, which references the image
• update screen shot in /README.md, which references the image