utils.py

import numpy as np
from PIL import Image
import scipy.ndimage
import insightface
import torch
import scipy

# Initialize InsightFace model
face_analyzer = insightface.app.FaceAnalysis(name='buffalo_l', providers=['CPUExecutionProvider'])
face_analyzer.prepare(ctx_id=0)


def image_grid(imgs, rows, cols):
    assert len(imgs) == rows*cols

    w, h = imgs[0].size
    grid = Image.new('RGB', size=(cols*w, rows*h))
    grid_w, grid_h = grid.size
    
    for i, img in enumerate(imgs):
        grid.paste(img, box=(i%cols*w, i//cols*h))
    return grid


def get_generator(seed, device):

    if seed is not None:
        if isinstance(seed, list):
            generator = [
                torch.Generator(device).manual_seed(seed_item) for seed_item in seed
            ]
        else:
            generator = torch.Generator(device).manual_seed(seed)
    else:
        generator = None

    return generator

def get_landmark_pil_insight(pil_image):
    """Get 68 facial landmarks using InsightFace."""
    img_np = np.array(pil_image.convert("RGB"))
    faces = face_analyzer.get(img_np)
    if not faces:
        return None
    landmarks = faces[0].kps  # shape: (5, 2) or (68, 2) depending on model

    if landmarks.shape[0] < 68:
        # InsightFace returns only 5 points: [left_eye, right_eye, nose, left_mouth, right_mouth]
        left_eye, right_eye, nose, left_mouth, right_mouth = landmarks
        # Approximate 68 landmarks (basic heuristic or fallback)
        return np.array([
            left_eye, right_eye, nose, left_mouth, right_mouth
        ])
    return landmarks

def align_face(pil_image):
    """Align a face from a PIL.Image, returning an aligned PIL.Image of size 512x512."""
    lm = get_landmark_pil_insight(pil_image)
    if lm is None or lm.shape[0] < 5:
        return pil_image

    eye_left, eye_right = lm[0], lm[1]
    eye_avg = (eye_left + eye_right) * 0.5
    eye_to_eye = eye_right - eye_left
    mouth_left, mouth_right = lm[3], lm[4]
    mouth_avg = (mouth_left + mouth_right) * 0.5
    eye_to_mouth = mouth_avg - eye_avg

    # The rest is your original alignment logic
    x = eye_to_eye - np.flipud(eye_to_mouth) * [-1, 1]
    x /= np.hypot(*x)
    x *= max(np.hypot(*eye_to_eye) * 2.0, np.hypot(*eye_to_mouth) * 1.8)
    y = np.flipud(x) * [-1, 1]
    c = eye_avg + eye_to_mouth * 0.1
    quad = np.stack([c - x - y, c - x + y, c + x + y, c + x - y])
    qsize = np.hypot(*x) * 2

    img = pil_image.convert("RGB")
    transform_size = 512
    output_size = 512
    enable_padding = True

    shrink = int(np.floor(qsize / output_size * 0.5))
    if shrink > 1:
        rsize = (int(np.rint(img.size[0] / shrink)), int(np.rint(img.size[1] / shrink)))
        img = img.resize(rsize, Image.Resampling.LANCZOS)
        quad /= shrink
        qsize /= shrink

    border = max(int(np.rint(qsize * 0.1)), 3)
    crop = (
        int(np.floor(min(quad[:, 0]))),
        int(np.floor(min(quad[:, 1]))),
        int(np.ceil(max(quad[:, 0]))),
        int(np.ceil(max(quad[:, 1])))
    )
    crop = (
        max(crop[0] - border, 0),
        max(crop[1] - border, 0),
        min(crop[2] + border, img.size[0]),
        min(crop[3] + border, img.size[1])
    )
    if crop[2] - crop[0] < img.size[0] or crop[3] - crop[1] < img.size[1]:
        img = img.crop(crop)
        quad -= crop[:2]

    pad = (
        int(np.floor(min(quad[:, 0]))),
        int(np.floor(min(quad[:, 1]))),
        int(np.ceil(max(quad[:, 0]))),
        int(np.ceil(max(quad[:, 1])))
    )
    pad = (
        max(-pad[0] + border, 0),
        max(-pad[1] + border, 0),
        max(pad[2] - img.size[0] + border, 0),
        max(pad[3] - img.size[1] + border, 0)
    )
    if enable_padding and max(pad) > border - 4:
        pad = np.maximum(pad, int(np.rint(qsize * 0.3)))
        img = np.pad(np.float32(img), ((pad[1], pad[3]), (pad[0], pad[2]), (0, 0)), 'reflect')
        h, w, _ = img.shape
        y, x, _ = np.ogrid[:h, :w, :1]
        mask = np.maximum(
            1.0 - np.minimum(np.float32(x) / pad[0], np.float32(w - 1 - x) / pad[2]),
            1.0 - np.minimum(np.float32(y) / pad[1], np.float32(h - 1 - y) / pad[3])
        )
        blur = qsize * 0.02
        img += (scipy.ndimage.gaussian_filter(img, [blur, blur, 0]) - img) * np.clip(mask * 3.0 + 1.0, 0.0, 1.0)
        img += (np.median(img, axis=(0, 1)) - img) * np.clip(mask, 0.0, 1.0)
        img = Image.fromarray(np.uint8(np.clip(np.rint(img), 0, 255)), 'RGB')
        quad += pad[:2]

    img = img.transform((transform_size, transform_size), Image.QUAD, (quad + 0.5).flatten(), Image.BILINEAR)
    if output_size < transform_size:
        img = img.resize((output_size, output_size), Image.Resampling.LANCZOS)

    return img