app.py

import os
import torch
import time
import threading
import json
import gc
from flask import Flask, request, jsonify, send_file, Response, stream_with_context
from werkzeug.utils import secure_filename
from PIL import Image
import io
import zipfile
import uuid
import traceback
from huggingface_hub import snapshot_download
from flask_cors import CORS
import numpy as np
import trimesh
from transformers import pipeline
from scipy.ndimage import gaussian_filter, uniform_filter

app = Flask(__name__)
CORS(app)  # Enable CORS for all routes

# Configure directories
UPLOAD_FOLDER = '/tmp/uploads'
RESULTS_FOLDER = '/tmp/results'
CACHE_DIR = '/tmp/huggingface'
ALLOWED_EXTENSIONS = {'png', 'jpg', 'jpeg'}

# Create necessary directories
os.makedirs(UPLOAD_FOLDER, exist_ok=True)
os.makedirs(RESULTS_FOLDER, exist_ok=True)
os.makedirs(CACHE_DIR, exist_ok=True)

# Set Hugging Face cache environment variables
os.environ['HF_HOME'] = CACHE_DIR
os.environ['TRANSFORMERS_CACHE'] = os.path.join(CACHE_DIR, 'transformers')
os.environ['HF_DATASETS_CACHE'] = os.path.join(CACHE_DIR, 'datasets')

app.config['UPLOAD_FOLDER'] = UPLOAD_FOLDER
app.config['MAX_CONTENT_LENGTH'] = 16 * 1024 * 1024  # 16MB max

# Job tracking dictionary
processing_jobs = {}

# Global model variables
depth_estimator = None
model_loaded = False
model_loading = False

# Configuration for processing
TIMEOUT_SECONDS = 240  # 4 minutes max for processing (increased for larger model)
MAX_DIMENSION = 512    # Max image dimension to process

# TimeoutError for handling timeouts
class TimeoutError(Exception):
    pass

# Thread-safe timeout implementation
def process_with_timeout(function, args, timeout):
    result = [None]
    error = [None]
    completed = [False]
    
    def target():
        try:
            result[0] = function(*args)
            completed[0] = True
        except Exception as e:
            error[0] = e
    
    thread = threading.Thread(target=target)
    thread.daemon = True
    thread.start()
    
    thread.join(timeout)
    
    if not completed[0]:
        if thread.is_alive():
            return None, TimeoutError(f"Processing timed out after {timeout} seconds")
        elif error[0]:
            return None, error[0]
    
    if error[0]:
        return None, error[0]
    
    return result[0], None

def allowed_file(filename):
    return '.' in filename and filename.rsplit('.', 1)[1].lower() in ALLOWED_EXTENSIONS

# Function to preprocess image
def preprocess_image(image_path):
    with Image.open(image_path) as img:
        img = img.convert("RGB")
        # Resize if the image is too large
        if img.width > MAX_DIMENSION or img.height > MAX_DIMENSION:
            # Calculate new dimensions while preserving aspect ratio
            if img.width > img.height:
                new_width = MAX_DIMENSION
                new_height = int(img.height * (MAX_DIMENSION / img.width))
            else:
                new_height = MAX_DIMENSION
                new_width = int(img.width * (MAX_DIMENSION / img.height))
            img = img.resize((new_width, new_height), Image.LANCZOS)
        
        return img

def load_model():
    global depth_estimator, model_loaded, model_loading
    
    if model_loaded:
        return depth_estimator
    
    if model_loading:
        # Wait for model to load if it's already in progress
        while model_loading and not model_loaded:
            time.sleep(0.5)
        return depth_estimator
    
    try:
        model_loading = True
        print("Starting model loading...")
        
        # Using DPT-Large which provides better detail than DPT-Hybrid
        model_name = "Intel/dpt-large"
        
        # Download model with retry mechanism
        max_retries = 3
        retry_delay = 5
        
        for attempt in range(max_retries):
            try:
                snapshot_download(
                    repo_id=model_name,
                    cache_dir=CACHE_DIR,
                    resume_download=True,
                )
                break
            except Exception as e:
                if attempt < max_retries - 1:
                    print(f"Download attempt {attempt+1} failed: {str(e)}. Retrying in {retry_delay} seconds...")
                    time.sleep(retry_delay)
                    retry_delay *= 2
                else:
                    raise
        
        # Initialize model with lower precision to save memory
        device = "cuda" if torch.cuda.is_available() else "cpu"
        
        # Load depth estimator pipeline
        depth_estimator = pipeline(
            "depth-estimation", 
            model=model_name,
            device=device if device == "cuda" else -1,
            cache_dir=CACHE_DIR
        )
        
        # Optimize memory usage
        if device == "cuda":
            torch.cuda.empty_cache()
        
        model_loaded = True
        print(f"Model loaded successfully on {device}")
        return depth_estimator
    
    except Exception as e:
        print(f"Error loading model: {str(e)}")
        print(traceback.format_exc())
        raise
    finally:
        model_loading = False

# Convert depth map to 3D mesh with enhanced detail
def depth_to_mesh(depth_map, image, resolution=100):
    """Convert depth map to 3D mesh with improved detail preservation"""
    # Convert depth_map to numpy array if it's a PIL Image
    if isinstance(depth_map, Image.Image):
        depth_map = np.array(depth_map)
    
    # Make sure the depth map is 2D
    if len(depth_map.shape) > 2:
        depth_map = np.mean(depth_map, axis=2) if depth_map.shape[2] > 1 else depth_map[:,:,0]
    
    # Apply bilateral filter to smooth the depth map while preserving edges
    # First, apply a slight gaussian filter to remove noise
    depth_map_smooth = gaussian_filter(depth_map, sigma=1.0)
    
    # Get dimensions
    h, w = depth_map_smooth.shape
    
    # Create a grid of points
    x = np.linspace(0, w-1, resolution)
    y = np.linspace(0, h-1, resolution)
    x_grid, y_grid = np.meshgrid(x, y)
    
    # Sample depth at grid points
    x_indices = x_grid.astype(int)
    y_indices = y_grid.astype(int)
    z_values = depth_map_smooth[y_indices, x_indices]
    
    # Normalize depth values with better scaling
    z_min, z_max = np.percentile(z_values, [2, 98])  # Removes outliers
    z_values = (z_values - z_min) / (z_max - z_min) if z_max > z_min else z_values
    z_values = z_values * 2.0  # Scale depth
    
    # Apply a local contrast enhancement to bring out details
    # Simple adaptive normalization
    window_size = resolution // 10
    if window_size > 0:
        local_mean = uniform_filter(z_values, size=window_size)
        local_var = uniform_filter(z_values**2, size=window_size) - local_mean**2
        local_std = np.sqrt(np.maximum(local_var, 0))
        
        # Enhance local contrast
        enhanced_z = (z_values - local_mean) / (local_std + 0.01) * 0.5 + z_values
        z_values = np.clip(enhanced_z, 0, None)  # Keep values positive
    
    # Normalize x and y coordinates
    x_grid = (x_grid / w - 0.5) * 2.0  # Map to -1 to 1
    y_grid = (y_grid / h - 0.5) * 2.0  # Map to -1 to 1
    
    # Create vertices
    vertices = np.vstack([x_grid.flatten(), -y_grid.flatten(), -z_values.flatten()]).T
    
    # Create faces (triangles)
    faces = []
    for i in range(resolution-1):
        for j in range(resolution-1):
            p1 = i * resolution + j
            p2 = i * resolution + (j + 1)
            p3 = (i + 1) * resolution + j
            p4 = (i + 1) * resolution + (j + 1)
            
            # Create two triangles for each grid cell
            faces.append([p1, p2, p4])
            faces.append([p1, p4, p3])
    
    faces = np.array(faces)
    
    # Create mesh
    mesh = trimesh.Trimesh(vertices=vertices, faces=faces)
    
    # Apply texturing if image is provided
    if image:
        # Convert to numpy array if needed
        if isinstance(image, Image.Image):
            img_array = np.array(image)
        else:
            img_array = image
            
        # Create simple texture by sampling the original image
        if resolution <= img_array.shape[0] and resolution <= img_array.shape[1]:
            # Create vertex colors by sampling the image
            vertex_colors = np.zeros((vertices.shape[0], 4), dtype=np.uint8)
            
            for i in range(resolution):
                for j in range(resolution):
                    img_x = min(int(j * img_array.shape[1] / resolution), img_array.shape[1]-1)
                    img_y = min(int(i * img_array.shape[0] / resolution), img_array.shape[0]-1)
                    
                    vertex_idx = i * resolution + j
                    if len(img_array.shape) == 3 and img_array.shape[2] == 3:  # RGB
                        vertex_colors[vertex_idx, :3] = img_array[img_y, img_x, :]
                        vertex_colors[vertex_idx, 3] = 255  # Alpha
                    elif len(img_array.shape) == 3 and img_array.shape[2] == 4:  # RGBA
                        vertex_colors[vertex_idx, :] = img_array[img_y, img_x, :]
                    else:
                        # Handle grayscale or other formats
                        gray_value = img_array[img_y, img_x]
                        vertex_colors[vertex_idx, :3] = [gray_value, gray_value, gray_value]
                        vertex_colors[vertex_idx, 3] = 255
            
            mesh.visual.vertex_colors = vertex_colors
    
    return mesh

@app.route('/health', methods=['GET'])
def health_check():
    return jsonify({
        "status": "healthy", 
        "model": "Depth-Based 3D Model Generator (DPT-Large)",
        "device": "cuda" if torch.cuda.is_available() else "cpu"
    }), 200

@app.route('/progress/<job_id>', methods=['GET'])
def progress(job_id):
    def generate():
        if job_id not in processing_jobs:
            yield f"data: {json.dumps({'error': 'Job not found'})}\n\n"
            return
            
        job = processing_jobs[job_id]
        
        # Send initial progress
        yield f"data: {json.dumps({'status': 'processing', 'progress': job['progress']})}\n\n"
        
        # Wait for job to complete or update
        last_progress = job['progress']
        check_count = 0
        while job['status'] == 'processing':
            if job['progress'] != last_progress:
                yield f"data: {json.dumps({'status': 'processing', 'progress': job['progress']})}\n\n"
                last_progress = job['progress']
            
            time.sleep(0.5)
            check_count += 1
            
            # If client hasn't received updates for a while, check if job is still running
            if check_count > 60:  # 30 seconds with no updates
                if 'thread_alive' in job and not job['thread_alive']():
                    job['status'] = 'error'
                    job['error'] = 'Processing thread died unexpectedly'
                    break
                check_count = 0
        
        # Send final status
        if job['status'] == 'completed':
            yield f"data: {json.dumps({'status': 'completed', 'progress': 100, 'result_url': job['result_url'], 'preview_url': job['preview_url']})}\n\n"
        else:
            yield f"data: {json.dumps({'status': 'error', 'error': job['error']})}\n\n"
    
    return Response(stream_with_context(generate()), mimetype='text/event-stream')

@app.route('/convert', methods=['POST'])
def convert_image_to_3d():
    # Check if image is in the request
    if 'image' not in request.files:
        return jsonify({"error": "No image provided"}), 400
    
    file = request.files['image']
    if file.filename == '':
        return jsonify({"error": "No image selected"}), 400
    
    if not allowed_file(file.filename):
        return jsonify({"error": f"File type not allowed. Supported types: {', '.join(ALLOWED_EXTENSIONS)}"}), 400
    
    # Get optional parameters with defaults
    try:
        mesh_resolution = min(int(request.form.get('mesh_resolution', 100)), 200)  # Limit max resolution
        output_format = request.form.get('output_format', 'obj').lower()
        detail_level = request.form.get('detail_level', 'medium').lower()  # New parameter for detail level
    except ValueError:
        return jsonify({"error": "Invalid parameter values"}), 400
    
    # Validate output format
    if output_format not in ['obj', 'glb']:
        return jsonify({"error": "Unsupported output format. Use 'obj' or 'glb'"}), 400
    
    # Adjust mesh resolution based on detail level
    if detail_level == 'high':
        mesh_resolution = min(mesh_resolution * 1.5, 200)
    elif detail_level == 'low':
        mesh_resolution = max(int(mesh_resolution * 0.7), 50)
    
    # Create a job ID
    job_id = str(uuid.uuid4())
    output_dir = os.path.join(RESULTS_FOLDER, job_id)
    os.makedirs(output_dir, exist_ok=True)
    
    # Save the uploaded file
    filename = secure_filename(file.filename)
    filepath = os.path.join(app.config['UPLOAD_FOLDER'], f"{job_id}_{filename}")
    file.save(filepath)
    
    # Initialize job tracking
    processing_jobs[job_id] = {
        'status': 'processing',
        'progress': 0,
        'result_url': None,
        'preview_url': None,
        'error': None,
        'output_format': output_format,
        'created_at': time.time()
    }
    
    # Start processing in a separate thread
    def process_image():
        thread = threading.current_thread()
        processing_jobs[job_id]['thread_alive'] = lambda: thread.is_alive()
        
        try:
            # Preprocess image
            processing_jobs[job_id]['progress'] = 5
            image = preprocess_image(filepath)
            processing_jobs[job_id]['progress'] = 10
            
            # Load model
            try:
                model = load_model()
                processing_jobs[job_id]['progress'] = 30
            except Exception as e:
                processing_jobs[job_id]['status'] = 'error'
                processing_jobs[job_id]['error'] = f"Error loading model: {str(e)}"
                return
            
            # Process image with thread-safe timeout
            try:
                def estimate_depth():
                    # Get depth map
                    result = model(image)
                    depth_map = result["depth"]
                    
                    # Convert to numpy array if needed
                    if isinstance(depth_map, torch.Tensor):
                        depth_map = depth_map.cpu().numpy()
                    elif hasattr(depth_map, 'numpy'):
                        depth_map = depth_map.numpy()
                    elif isinstance(depth_map, Image.Image):
                        depth_map = np.array(depth_map)
                    
                    return depth_map
                
                depth_map, error = process_with_timeout(estimate_depth, [], TIMEOUT_SECONDS)
                
                if error:
                    if isinstance(error, TimeoutError):
                        processing_jobs[job_id]['status'] = 'error'
                        processing_jobs[job_id]['error'] = f"Processing timed out after {TIMEOUT_SECONDS} seconds"
                        return
                    else:
                        raise error
                        
                processing_jobs[job_id]['progress'] = 60
                
                # Create mesh from depth map
                mesh_resolution_int = int(mesh_resolution)
                mesh = depth_to_mesh(depth_map, image, resolution=mesh_resolution_int)
                processing_jobs[job_id]['progress'] = 80
                
            except Exception as e:
                error_details = traceback.format_exc()
                processing_jobs[job_id]['status'] = 'error'
                processing_jobs[job_id]['error'] = f"Error during processing: {str(e)}"
                print(f"Error processing job {job_id}: {str(e)}")
                print(error_details)
                return
            
            # Export based on requested format
            try:
                if output_format == 'obj':
                    obj_path = os.path.join(output_dir, "model.obj")
                    mesh.export(obj_path, file_type='obj')
                    
                    # Create a zip file with OBJ and MTL
                    zip_path = os.path.join(output_dir, "model.zip")
                    with zipfile.ZipFile(zip_path, 'w') as zipf:
                        zipf.write(obj_path, arcname="model.obj")
                        mtl_path = os.path.join(output_dir, "model.mtl")
                        if os.path.exists(mtl_path):
                            zipf.write(mtl_path, arcname="model.mtl")
                    
                    processing_jobs[job_id]['result_url'] = f"/download/{job_id}"
                    processing_jobs[job_id]['preview_url'] = f"/preview/{job_id}"
                    
                elif output_format == 'glb':
                    # Export as GLB
                    glb_path = os.path.join(output_dir, "model.glb")
                    mesh.export(glb_path, file_type='glb')
                    
                    processing_jobs[job_id]['result_url'] = f"/download/{job_id}"
                    processing_jobs[job_id]['preview_url'] = f"/preview/{job_id}"
                
                # Update job status
                processing_jobs[job_id]['status'] = 'completed'
                processing_jobs[job_id]['progress'] = 100
                print(f"Job {job_id} completed successfully")
            except Exception as e:
                error_details = traceback.format_exc()
                processing_jobs[job_id]['status'] = 'error'
                processing_jobs[job_id]['error'] = f"Error exporting model: {str(e)}"
                print(f"Error exporting model for job {job_id}: {str(e)}")
                print(error_details)
            
            # Clean up temporary file
            if os.path.exists(filepath):
                os.remove(filepath)
            
            # Force garbage collection to free memory
            gc.collect()
            if torch.cuda.is_available():
                torch.cuda.empty_cache()
            
        except Exception as e:
            # Handle errors
            error_details = traceback.format_exc()
            processing_jobs[job_id]['status'] = 'error'
            processing_jobs[job_id]['error'] = f"{str(e)}\n{error_details}"
            print(f"Error processing job {job_id}: {str(e)}")
            print(error_details)
            
            # Clean up on error
            if os.path.exists(filepath):
                os.remove(filepath)
    
    # Start processing thread
    processing_thread = threading.Thread(target=process_image)
    processing_thread.daemon = True
    processing_thread.start()
    
    # Return job ID immediately
    return jsonify({"job_id": job_id}), 202  # 202 Accepted

@app.route('/download/<job_id>', methods=['GET'])
def download_model(job_id):
    if job_id not in processing_jobs or processing_jobs[job_id]['status'] != 'completed':
        return jsonify({"error": "Model not found or processing not complete"}), 404
    
    # Get the output directory for this job
    output_dir = os.path.join(RESULTS_FOLDER, job_id)
    
    # Determine file format from the job data
    output_format = processing_jobs[job_id].get('output_format', 'obj')
    
    if output_format == 'obj':
        zip_path = os.path.join(output_dir, "model.zip")
        if os.path.exists(zip_path):
            return send_file(zip_path, as_attachment=True, download_name="model.zip")
    else:  # glb
        glb_path = os.path.join(output_dir, "model.glb")
        if os.path.exists(glb_path):
            return send_file(glb_path, as_attachment=True, download_name="model.glb")
    
    return jsonify({"error": "File not found"}), 404

@app.route('/preview/<job_id>', methods=['GET'])
def preview_model(job_id):
    if job_id not in processing_jobs or processing_jobs[job_id]['status'] != 'completed':
        return jsonify({"error": "Model not found or processing not complete"}), 404
    
    # Get the output directory for this job
    output_dir = os.path.join(RESULTS_FOLDER, job_id)
    output_format = processing_jobs[job_id].get('output_format', 'obj')

    if output_format == 'obj':
        obj_path = os.path.join(output_dir, "model.obj")
        if os.path.exists(obj_path):
            return send_file(obj_path, mimetype='model/obj')
    else:  # glb
        glb_path = os.path.join(output_dir, "model.glb")
        if os.path.exists(glb_path):
            return send_file(glb_path, mimetype='model/gltf-binary')
    
    return jsonify({"error": "Model file not found"}), 404

# Cleanup old jobs periodically
def cleanup_old_jobs():
    current_time = time.time()
    job_ids_to_remove = []
    
    for job_id, job_data in processing_jobs.items():
        # Remove completed jobs after 1 hour
        if job_data['status'] == 'completed' and (current_time - job_data.get('created_at', 0)) > 3600:
            job_ids_to_remove.append(job_id)
        # Remove error jobs after 30 minutes
        elif job_data['status'] == 'error' and (current_time - job_data.get('created_at', 0)) > 1800:
            job_ids_to_remove.append(job_id)
    
    # Remove the jobs
    for job_id in job_ids_to_remove:
        output_dir = os.path.join(RESULTS_FOLDER, job_id)
        try:
            import shutil
            if os.path.exists(output_dir):
                shutil.rmtree(output_dir)
        except Exception as e:
            print(f"Error cleaning up job {job_id}: {str(e)}")
        
        # Remove from tracking dictionary
        if job_id in processing_jobs:
            del processing_jobs[job_id]
    
    # Schedule the next cleanup
    threading.Timer(300, cleanup_old_jobs).start()  # Run every 5 minutes

@app.route('/', methods=['GET'])
def index():
    return jsonify({
        "message": "Image to 3D API is running (DPT-Large Model)", 
        "endpoints": ["/convert", "/progress/<job_id>", "/download/<job_id>", "/preview/<job_id>"],
        "parameters": {
            "mesh_resolution": "Integer (50-200), controls mesh density",
            "output_format": "obj or glb",
            "detail_level": "low, medium, or high"
        }
    }), 200

if __name__ == '__main__':
    # Start the cleanup thread
    cleanup_old_jobs()
    
    # Use port 7860 which is standard for Hugging Face Spaces
    port = int(os.environ.get('PORT', 7860))
    app.run(host='0.0.0.0', port=port)