app.py

# app.py
import os, glob, traceback
import numpy as np
from PIL import Image
import gradio as gr

# Use Keras 3 with the TensorFlow backend
os.environ.setdefault("KERAS_BACKEND", "tensorflow")
import keras  # Keras 3.x
from huggingface_hub import hf_hub_download

HF_MODEL_ID = "Vedag812/xray_cnn"
CLASS_NAMES = ["NORMAL", "PNEUMONIA"]


def load_model():
    model_path = hf_hub_download(HF_MODEL_ID, filename="xray_cnn.keras")
    # safe_mode=False allows loading models saved with older or custom configs
    model = keras.saving.load_model(model_path, compile=False, safe_mode=False)
    return model

def _infer_input_shape(model):
    # returns (H, W, C)
    try:
        shp = tuple(model.inputs[0].shape)
    except Exception:
        shp = getattr(model, "input_shape", None)
        if shp is None:
            return 150, 150, 1
    if len(shp) < 4:
        return 150, 150, 1
    H = int(shp[1]) if shp[1] is not None else 150
    W = int(shp[2]) if shp[2] is not None else 150
    C = int(shp[3]) if shp[3] is not None else 1
    return H, W, C

def preprocess(pil_img: Image.Image, target):
    H, W, C = target
    g = pil_img.convert("L").resize((W, H))
    arr = np.array(g).astype("float32") / 255.0  # (H, W)
    if C == 1:
        x = np.expand_dims(arr, axis=(0, -1))           # (1,H,W,1)
    elif C == 3:
        x = np.expand_dims(np.stack([arr]*3, axis=-1), 0)  # (1,H,W,3)
    else:
        x = np.expand_dims(np.repeat(arr[..., None], C, axis=-1), 0)
    return x

def predict_fn(pil_img: Image.Image):
    try:
        if pil_img is None:
            return {"NORMAL": 0.0, "PNEUMONIA": 0.0}, "Please upload an image or pick a sample."
        model = load_model()
        H, W, C = _infer_input_shape(model)
        x = preprocess(pil_img, (H, W, C))
        y = model.predict(x, verbose=0)
        prob = float(np.ravel(y)[0])  # sigmoid output
        idx = int(prob > 0.5)
        conf = prob if idx == 1 else 1 - prob
        probs = {CLASS_NAMES[0]: 1 - prob, CLASS_NAMES[1]: prob}
        msg = f"Prediction: {CLASS_NAMES[idx]} | Confidence: {conf*100:.2f}%"
        return probs, msg
    except Exception as e:
        tip = (
            "If this persists, make sure the Space has keras>=3 and tensorflow>=2.16."
        )
        err = f"⚠️ Error during prediction:\n\n{e}\n\n{tip}"
        # Optional: uncomment next line to print full stack to the Space logs
        # print(traceback.format_exc())
        return {"NORMAL": 0.0, "PNEUMONIA": 0.0}, err

def list_examples():
    files = []
    for pat in ("images/*.jpeg", "images/*.jpg", "images/*.png"):
        files.extend(glob.glob(pat))
    return [[p] for p in sorted(files)]

with gr.Blocks(css="""
.gradio-container {max-width: 980px !important; margin: auto;}
#title {text-align:center;}
""") as demo:
    gr.Markdown("<h1 id='title'>Chest X-Ray Classification</h1>")
    gr.Markdown("Upload an image or click a sample. The model predicts NORMAL or PNEUMONIA.")

    with gr.Row():
        with gr.Column(scale=2):
            inp = gr.Image(type="pil", image_mode="L", label="Upload X-ray")
            with gr.Row():
                btn = gr.Button("Predict", variant="primary")
                gr.ClearButton(components=[inp])
            gr.Markdown("### Samples")
            gr.Examples(examples=list_examples(), inputs=inp, examples_per_page=12)
        with gr.Column(scale=1):
            probs = gr.Label(num_top_classes=2, label="Class probabilities")
            out_text = gr.Markdown()

    btn.click(predict_fn, inputs=inp, outputs=[probs, out_text])
    inp.change(predict_fn, inputs=inp, outputs=[probs, out_text])

if __name__ == "__main__":
    demo.launch()