Duplicate from Xian1057/histopathological_img_cls

Co-authored-by: Kee ChiongHean <Xian1057@users.noreply.huggingface.co>

.gitattributes

@@ -0,0 +1,34 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,14 @@
+---
+title: Histopath Cls
+emoji: 🐨
+colorFrom: indigo
+colorTo: pink
+sdk: streamlit
+sdk_version: 1.21.0
+app_file: app.py
+pinned: false
+license: mit
+duplicated_from: Xian1057/histopathological_img_cls
+---
+
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,116 @@
+from fastai.vision.all import *
+from io import BytesIO
+from PIL import Image
+from utils.cam import grad_cam
+from utils.tfms import AlbTransform, get_augs
+import requests
+import streamlit as st
+
+
+"""
+# Histopathologic Cancer Detection
+## Is this an cancerous cell?
+this is a web app to predict whether a cell is cancerous or not.
+"""
+
+# List of example image image paths
+images = {
+    "Image 1": "example/c2a8ef295e49b8012e5bc83917a305057eaa1932.tif",
+    "Image 2": "example/87aed2dd32c7fe17c3af56878abd62eb5f37f925.tif",
+    "Image 3": "example/4ba0d1c62230781533c2df1d95e5b4c8f5b650a3.tif",
+    "Image 4": "example/a24ce148f6ffa7ef8eefb4efb12ebffe8dd700da.tif",
+    "Image 5": "example/d2dd0de8e583a5475a07c2f92fa3f06c7fabcd42.tif",
+    # add more images as needed
+}
+
+
+def predict(learn, img):
+    img = PILImage.create(img)
+    pred, key, probs = learn.predict(img)
+    grad_img = grad_cam(learn, img, pred)
+    
+    col1, col2 = st.columns(2)
+    with col1:
+        st.image(img, caption='Original Image', use_column_width=True)
+    
+    with col2:
+        st.image(grad_img, caption='Grad-CAM Image', use_column_width=True)
+        
+    # st.write(learn_inf.predict(img))
+
+    f"""
+    ## This **{'is ' if pred == '1' else 'is not'}** an cancerous cell.
+    ### Rediction result: {pred}
+    ### Probability of {pred}: {probs[key].item()*100: .2f}%
+    """
+
+
+path = "./"
+learn_inf = load_learner(path + "resnet50.pkl")
+
+option = st.radio("", ["Upload Image", "Image URL", "Download Example Image", "Example"])
+
+if option == "Upload Image":
+    uploaded_file = st.file_uploader("Please upload an image.")
+
+    if uploaded_file is not None:
+        predict(learn_inf, uploaded_file)
+
+elif option == "Image URL":
+    url = st.text_input("Please input a url.")
+
+    if url != "":
+        try:
+            response = requests.get(url)
+            pil_img = PILImage.create(BytesIO(response.content))
+            predict(learn_inf, pil_img)
+
+        except:
+            st.text("Problem reading image from", url)
+elif option == "Download Example Image":
+    # Create columns for the images
+    cols = st.columns(len(images))
+    
+    
+    # Initialize selected_image in session state if it doesn't exist
+    if 'selected_image' not in st.session_state:
+        st.session_state['selected_image'] = None
+    
+    for i, (img_name, img_path) in enumerate(images.items()):
+        # Open the image file
+        img = Image.open(img_path)
+    
+        # Display the image in a column
+        cols[i].image(img, caption=img_name, use_column_width=True)
+    
+        # Create a button for selecting the image
+        if cols[i].button(f"Select {img_name}", key=img_name):
+            st.session_state['selected_image'] = img_path
+    
+    # If an image has been selected, show the download button
+    if st.session_state['selected_image'] is not None:
+        # Open the selected image file
+        img = Image.open(st.session_state['selected_image'])
+    
+        # Convert the image to a byte array
+        img_byte_arr = io.BytesIO()
+        img.save(img_byte_arr, format='PNG')
+        img_byte_arr = img_byte_arr.getvalue()
+    
+        # Create the download button for the selected image
+        st.download_button(
+            label="Download selected image",
+            data=img_byte_arr,
+            file_name="selected_image.png",
+            mime="image/png",
+        ) 
+else:
+    option_1 = st.selectbox(
+    'Example Prediction',
+    ('Positive Image', 'Negative Image'))
+
+    if option_1 == 'Positive Image':
+        predict(learn_inf, img='example/c2a8ef295e49b8012e5bc83917a305057eaa1932.tif')
+    elif option_1 == 'Negative Image':
+        predict(learn_inf, img='example/87aed2dd32c7fe17c3af56878abd62eb5f37f925.tif')
+            

requirements.txt

@@ -0,0 +1,5 @@
+fastbook
+timm
+albumentations
+grad-cam
+altair<5

resnet50.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ba8099d08b3224eab0cf7e48c98a02fdee3cb3cf6bc9449568275322cc04986
+size 98319417

utils/cam.py

@@ -0,0 +1,26 @@
+import numpy as np
+import cv2
+
+from pytorch_grad_cam import GradCAM
+from pytorch_grad_cam.utils.model_targets import ClassifierOutputTarget
+from pytorch_grad_cam.utils.image import show_cam_on_image, preprocess_image
+from PIL import Image
+
+def grad_cam(model, img, cls):
+    model.eval()
+    img = np.array(img)
+    img = cv2.resize(img, (224, 224))
+    img = np.float32(img) / 255
+    input_tensor = preprocess_image(img, mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
+
+    # The target for the CAM is the Bear category.
+    # As usual for classication, the target is the logit output
+    # before softmax, for that category.
+    targets = [ClassifierOutputTarget(1)]
+    target_layers = [model.layer4[-1]]
+    with GradCAM(model=model, target_layers=target_layers) as cam:
+        grayscale_cams = cam(input_tensor=input_tensor, targets=targets)
+        cam_image = show_cam_on_image(img, grayscale_cams[0, :], use_rgb=True)
+    cam = np.uint8(255*grayscale_cams[0, :])
+    cam = cv2.merge([cam, cam, cam])
+    return Image.fromarray(cam_image)

utils/tfms.py

@@ -0,0 +1,18 @@
+from fastai.vision.all import *
+import albumentations as Alb
+
+class AlbTransform(Transform):
+    def __init__(self, aug): self.aug = aug
+    def encodes(self, img: PILImage):
+        aug_img = self.aug(image=np.array(img))['image']
+        return PILImage.create(aug_img)
+    
+def get_augs(): return Alb.Compose([
+    Alb.Transpose(),
+    Alb.Flip(),
+    Alb.RandomRotate90(),
+    Alb.HueSaturationValue(
+      hue_shift_limit=5, 
+      sat_shift_limit=5, 
+      val_shift_limit=5 ),
+])