🙋‍♂️ This Project is divided across three repositories, This repo deals with the app/model deployed on cloud

• For Frontend App see Frontend Repo
• For Model Training see Training Notebook

What is this? :

• The Project aims to solve the problem of identification of medicinal herbs.
• The Machine Learning Model uses ResNet, with a validation accuracy of 98% and testing accuracy of 96%.
• The training and testing dataset contains over 1500 images across 30 medicinal herb species.

Tech Stack Used :

• Python 3
• Flask
• NumPy
• Pandas
• PyTorch
• Tensorflow-Keras
• FineTuned ResNet50 model trained on Mendeley Medicinal Leaf Dataset
• Dataset: S, Roopashree; J, Anitha (2020), “Medicinal Leaf Dataset”, Mendeley Data, V1, doi: 10.17632/nnytj2v3n5.1
  
• Google Cloud
• GitHub

Model Deployment as a Flask WebApp on Google Cloud Services ☁️ :

Gcloud SDK shell commands to push, and deploy the service (should've set up GCloud beforehand, you don't need Docker on your machine though):

• if you are running the shell in the same directory as your Dockerfile (which you probably should), replace Path/to/Dockerfile with . (a period or fullstop)

• replace projectid with the ID associated with your project on the Gcloud console.

• replace function with the name of the function you want your POST request from frontend to hit On.

   gcloud builds submit --tag gcr.io/{projectid}/{function} Path/to/Dockerfile
  

   gcloud run deploy --image gcr.io/{projectid}/{function}  --platform managed
  

Example Post request :

    import requests
    response = requests.post("YourServiceURLfromGcloud", files={'file': open("imgPath", 'rb')})

For instance, a post request for my deployed service on GCP is the following :

  response = requests.post("https://aushadhub-prcsxigeha-el.a.run.app/upload", files={'file': open("imgPath", 'rb')})

• if you're encountering 403 Error (proxy or tunnel connection error), make sure you are not using Python-Anywhere to host your front-end, else try adding header and/or proxy to your requests :

  header = { Define your header, to find it just run "navigator.userAgent" on your Chrome dev console }
  proxy = { define your proxy, you may find it on the internet }
  response = requests.post("https://aushadhub-prcsxigeha-el.a.run.app/upload", files={'file': open("imgPath", 'rb')}, header=header, proxies=proxies)
  

Return Type (JSON):

{ "prediction": "predName", "info": "description", "confidence_level": "probability" }

Example Code for unpacking returned JSON :

      import json
      responsedata = response.json()

      prediction = responsedata['prediction']
      confidence_level = responsedata['confidence_level']
      info = responsedata['info']

In Action :

TestPost.mp4