inference_classifier.py

import cv2
import numpy as np
import streamlit as st
from PIL import Image
import os

# ✅ Updated Mediapipe import
from mediapipe import solutions as mp_solutions

from isl_utils import load_model, extract_landmarks, predict

MODEL_FILE = "./model.p"

st.title("✋ ISL Gesture → Alphabet Recognition")

if not os.path.exists(MODEL_FILE):
    st.error("❌ No trained model found. Please run train_classifier.py first.")
    st.stop()

model = load_model(MODEL_FILE)

st.write("📷 Show a hand gesture to the camera below:")

img_file = st.camera_input("Capture gesture")

if img_file is not None:
    image = Image.open(img_file)
    frame = cv2.cvtColor(np.array(image), cv2.COLOR_RGB2BGR)
    rgb_img = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)

    # ✅ Updated mediapipe usage
    with mp_solutions.hands.Hands(
        static_image_mode=True,
        min_detection_confidence=0.3
    ) as hands_detector:
        results = hands_detector.process(rgb_img)

        if results.multi_hand_landmarks:
            hand_landmarks = results.multi_hand_landmarks[0]
            landmarks = extract_landmarks(hand_landmarks)
            predicted_letter = predict(model, landmarks)
            st.success(f"✅ Predicted gesture: **{predicted_letter}**")
        else:
            st.warning("⚠️ No hand detected. Try again.")