Upload app.py

app.py

@@ -1,18 +1,17 @@
 from io import BytesIO
 import gradio as gr
 import tensorflow as tf
+import logging
 import requests
 import numpy as np
 from PIL import Image
 from tensorflow.keras.utils import CustomObjectScope
 from tensorflow.keras.layers.experimental.preprocessing import RandomHeight
 
-import os
-from fastapi import FastAPI
-from fastapi.responses import JSONResponse
-
-app = FastAPI()
-
+logging.basicConfig(
+    level=logging.WARNING,  
+    format="%(asctime)s - %(levelname)s - %(message)s",  
+)
 with CustomObjectScope({'RandomHeight': RandomHeight}):
     model_0 = tf.keras.models.load_model('bestmodel.h5')
 
@@ -29,25 +28,37 @@ if gpus:
     except RuntimeError as e:
         print(e)
 
+def clean_url(input_string):
+    if "http" in input_string:
+        return input_string[input_string.find("http"):]
+    return input_string
+
 
 def classify_image(inp):
-    Warning(inp)
     # Convert to PIL Image if we have a numpy array
     image = None
+    logging.warning("entree dans ckassify_image")
     if isinstance(inp, np.ndarray):
         image = Image.fromarray(inp)
-        Warning("Image is a numpy array")
-        Warning(image)
+        logging.warning("1")
+
+    if isinstance(inp, str) :
+        logging.warning("2")
+
+        inp = clean_url(inp)
+        response = requests.get(inp)
+        response.raise_for_status()
+        image = Image.open(BytesIO(response.content))
     if isinstance(inp, str) and (inp.startswith("http://") or inp.startswith("https://")):
+        logging.warning("3")
         response = requests.get(inp)
         response.raise_for_status()
         image = Image.open(BytesIO(response.content))
-        Warning("Image is url")
-        Warning(image)
+       
     else:
-        Warning("Image is ???")
-        Warning(image)
-        image = inp
+              logging.warning("4")
+
+              image = inp
     
     # Resize image to 224x224
     image = image.resize((224, 224), Image.Resampling.LANCZOS)
@@ -81,12 +92,6 @@ def classify_image(inp):
     return output
 
 
-@app.get("/predict2")
-async def predict_custom_controller(image_url: str):
-    return JSONResponse(content=classify_image(image_url))
-
-
-
 gr.Interface(
     fn=classify_image, inputs=gr.Image(type="filepath",label="Image ou Url"), outputs="text",live=True,title="API de détection des images violentes",
 ).launch(share=True)