Upload 4 files

app.py

@@ -0,0 +1,88 @@
+
+import gradio as gr
+import cv2
+import mediapipe as mp
+import numpy as np
+import json
+import time
+from gtts import gTTS
+import tempfile
+import requests
+from PIL import Image
+from io import BytesIO
+
+with open("gestures_rules.json", "r") as f:
+    gesture_data = json.load(f)["gestures"]
+
+mp_hands = mp.solutions.hands
+hands = mp_hands.Hands(max_num_hands=1, min_detection_confidence=0.7, min_tracking_confidence=0.7)
+
+def get_finger_states(hand_landmarks):
+    tips = [4, 8, 12, 16, 20]
+    pips = [2, 6, 10, 14, 18]
+    return [1 if hand_landmarks.landmark[t].y < hand_landmarks.landmark[p].y else 0 for t, p in zip(tips, pips)]
+
+def detect_gesture(states):
+    for name, rule in gesture_data.items():
+        if rule["pattern"] == states:
+            return name
+    return None
+
+def speak_text(text):
+    tts = gTTS(text=text)
+    f = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3")
+    tts.save(f.name)
+    return f.name
+
+def process(frame, sentence, last_char, last_time):
+    if frame is None:
+        return frame, sentence, last_char, last_time
+    frame = cv2.flip(frame, 1)
+    rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
+    res = hands.process(rgb)
+    if res.multi_hand_landmarks:
+        states = get_finger_states(res.multi_hand_landmarks[0])
+        char = detect_gesture(states)
+        now = time.time()
+        if char and char != last_char and now - last_time > 1 and len(char) == 1:
+            sentence += char
+            last_char = char
+            last_time = now
+    return frame, sentence, last_char, last_time
+
+def speak(sentence):
+    return speak_text(sentence) if sentence.strip() else None
+
+def clear():
+    return "", "", 0.0
+
+logo = Image.open(BytesIO(requests.get(
+    "https://raw.githubusercontent.com/imarshbir/Hand2Voice/main/logo.png"
+).content))
+
+with open("styles.css") as f:
+    css = f.read()
+
+with gr.Blocks(css=css, title="Hand2Voice") as demo:
+    gr.Image(logo, show_label=False, height=100)
+    gr.HTML("<div class='mission-box'><div class='mission-title'>Bridging Communication Gaps with AI</div><div class='mission-text'>Sign language to speech in real time.</div></div>")
+
+    with gr.Row():
+        cam = gr.Image(source="webcam", streaming=True, type="numpy")
+        out = gr.HTML("<div class='output-text'>🤚 Waiting for gestures...</div>")
+
+    speak_btn = gr.Button("🔊 Speak Sentence")
+    clear_btn = gr.Button("🧹 Clear Text")
+    audio = gr.Audio(autoplay=True)
+
+    sentence = gr.State("")
+    last_char = gr.State("")
+    last_time = gr.State(0.0)
+
+    cam.stream(process, [cam, sentence, last_char, last_time], [cam, sentence, last_char, last_time])\
+        .then(lambda s: f"<div class='output-text'>{s}</div>", sentence, out)
+
+    speak_btn.click(speak, sentence, audio)
+    clear_btn.click(clear, outputs=[sentence, last_char, last_time])
+
+demo.launch()

gestures_rules.json

@@ -0,0 +1,31 @@
+
+{
+  "gestures": {
+    "A": { "pattern": [0,0,0,0,0] },
+    "B": { "pattern": [0,1,1,1,1] },
+    "C": { "pattern": [1,1,1,1,0] },
+    "D": { "pattern": [0,1,0,0,0] },
+    "E": { "pattern": [0,0,0,0,1] },
+    "F": { "pattern": [1,0,1,1,1] },
+    "G": { "pattern": [1,1,0,0,0] },
+    "H": { "pattern": [0,1,1,0,0] },
+    "I": { "pattern": [0,0,0,0,1] },
+    "J": { "pattern": [1,0,0,0,1] },
+    "K": { "pattern": [1,1,1,0,0] },
+    "L": { "pattern": [1,1,0,0,0] },
+    "M": { "pattern": [0,1,1,1,0] },
+    "N": { "pattern": [0,1,1,0,0] },
+    "O": { "pattern": [1,1,1,1,1] },
+    "P": { "pattern": [1,1,0,1,0] },
+    "Q": { "pattern": [1,1,0,0,1] },
+    "R": { "pattern": [0,1,1,0,1] },
+    "S": { "pattern": [0,0,0,0,0] },
+    "T": { "pattern": [1,0,0,0,0] },
+    "U": { "pattern": [0,1,1,0,0] },
+    "V": { "pattern": [0,1,1,0,1] },
+    "W": { "pattern": [0,1,1,1,0] },
+    "X": { "pattern": [0,1,0,0,1] },
+    "Y": { "pattern": [1,0,0,0,1] },
+    "Z": { "pattern": [1,1,0,0,0] }
+  }
+}

requirements.txt

@@ -0,0 +1,7 @@
+gradio
+opencv-python-headless
+mediapipe
+numpy
+gTTS
+Pillow
+requests

styles.css

@@ -0,0 +1 @@
+/* Custom CSS from user (unchanged) */