Update app.py

app.py

@@ -2,23 +2,24 @@ import streamlit as st
 import requests
 import os
 from PIL import Image
-from transformers import AutoImageProcessor, AutoModelForImageClassification
 import torch
 import cv2
 import time
+import numpy as np
+from tensorflow.keras.models import load_model
 
-# Load the open-source Hugging Face sign language model
-processor = AutoImageProcessor.from_pretrained("microsoft/resnet-50")
-model = AutoModelForImageClassification.from_pretrained("microsoft/resnet-50")
+# Load the DeepASL model for live ASL alphabet classification
+MODEL_PATH = "asl_alphabet_model.h5"
+model = load_model(MODEL_PATH)
 
-# Function for sign classification
-def classify_sign(image):
-    image = image.convert("RGB")
-    inputs = processor(images=image, return_tensors="pt")
-    outputs = model(**inputs)
-    prediction = torch.argmax(outputs.logits, dim=-1).item()
-    labels = ["Hello", "Thank You", "Yes", "No", "Please"]  # Update with actual model labels if available
-    return labels[prediction % len(labels)]
+# Function for ASL classification
+def classify_asl(image):
+    image = image.resize((64, 64))  # Resize image to model input size
+    image = np.array(image) / 255.0  # Normalize
+    image = np.expand_dims(image, axis=0)  # Add batch dimension
+    prediction = model.predict(image)
+    labels = list("ABCDEFGHIJKLMNOPQRSTUVWXYZ")  # ASL alphabet labels
+    return labels[np.argmax(prediction)]
 
 # Streamlit UI
 def main():
@@ -34,15 +35,15 @@ def main():
             st.button("📞 Contact Us", use_container_width=True)
             st.button("💬 Feedback", use_container_width=True)
 
-    tab1, tab2, tab3, tab4 = st.tabs(["Image Load", "Take Picture", "Live", "Text2Sign"])
+    tab1, tab2, tab3, tab4 = st.tabs(["Image Load", "Take Picture", "Live ASL", "Text2Sign"])
 
     with tab1:
         st.subheader("📸 Image Load")
-        uploaded_image = st.file_uploader("Upload an image of a hand gesture", type=["png", "jpg", "jpeg"])
+        uploaded_image = st.file_uploader("Upload an image of an ASL alphabet gesture", type=["png", "jpg", "jpeg"])
         if uploaded_image:
             image = Image.open(uploaded_image)
             st.image(image, caption="Uploaded Image", use_container_width=True)
-            gesture = classify_sign(image)
+            gesture = classify_asl(image)
             st.success(f"Detected Gesture: {gesture}")
 
     with tab2:
@@ -51,11 +52,11 @@ def main():
         if camera_image:
             image = Image.open(camera_image)
             st.image(image, caption="Captured Image", use_container_width=True)
-            gesture = classify_sign(image)
+            gesture = classify_asl(image)
             st.success(f"Detected Gesture: {gesture}")
 
     with tab3:
-        st.subheader("📹 Live")
+        st.subheader("📹 Live ASL")
         if st.button("Enable Cam"):
             cap = cv2.VideoCapture(0)
             stframe = st.image([])
@@ -65,10 +66,10 @@ def main():
                 if not ret:
                     break
                 image = Image.fromarray(cv2.cvtColor(frame, cv2.COLOR_BGR2RGB))
-                gesture = classify_sign(image)
+                gesture = classify_asl(image)
                 frame = cv2.putText(frame, f"Gesture: {gesture}", (10, 50), cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 255, 0), 2)
                 stframe.image(frame, channels="BGR", use_container_width=True)
-                time.sleep(5)
+                time.sleep(1)
             cap.release()
 
     with tab4: