Create app.py

app.py

@@ -0,0 +1,48 @@
+# -*- coding: utf-8 -*-
+"""
+Created on Sun Dec 25 08:38:00 2022
+
+@author: ROSHAN
+"""
+
+import tensorflow as tf
+import gradio as gr
+import numpy as np
+import cv2
+from PIL import Image as im
+from matplotlib import pyplot as plt
+cls=['Angry', 'Disgust', 'Fear', 'Happy', 'Sad', 'Surprise', 'Neutral']
+model = tf.keras.models.load_model("D:/56fer.h5")
+face_cascade = cv2.CascadeClassifier('D:/haarcascade_frontalface_default.xml')
+def show(img):
+    img=img[:, :, ::-1].copy() 
+    
+    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
+    faces = face_cascade.detectMultiScale(gray, 1.5, 1)
+    r=[]
+    x=faces[0][0]
+    y=faces[0][1]
+    w=faces[0][2]
+    h=faces[0][3]
+    cv2.rectangle(img,(x,y),(x+w,y+h),(255,255,0),2) 
+    r.append(img)
+    sharp_kernel = np.array([[0, -1, 0],
+                [-1, 5, -1],
+                [0, -1, 0]])
+    sharp_img = cv2.filter2D(src=gray, ddepth=-1, kernel=sharp_kernel)
+    
+    crop_img = sharp_img[y:y+h, x:x+w]
+    
+    npa=np.array(crop_img)/255.0
+    predictions = model.predict(np.resize(npa,(48,48)).reshape(-1,48,48,1))
+    score =predictions[0]
+    score=tf.nn.softmax(predictions[0])
+    plt.figure()
+    confidences = {cls[i]: float(score[i]) for i in range(len(cls))}   
+    return  confidences
+demo = gr.Interface(
+    fn=show,
+    inputs="image",
+    outputs=gr.outputs.Label(num_top_classes=7),
+)
+demo.launch()