from argparse import ArgumentParser import time import numpy as np import PIL import PIL.Image import os import scipy import scipy.ndimage import insightface import multiprocessing as mp import math def get_landmark(filepath, face_detector): """get landmark with InsightFace :return: np.array shape=(68, 2) """ if isinstance(filepath, str): img = PIL.Image.open(filepath) img = np.array(img) else: img = filepath faces = face_detector.get(img) if len(faces) == 0: print('Error: no face detected!') return None # Assume the first detected face is the target face = faces[0] lm = face.landmark_2d_106[:, :2] # Use 106-point landmarks return lm def align_face(filepath, face_detector): """ :param filepath: str :return: PIL Image """ lm = get_landmark(filepath, face_detector) if lm is None: return None # Use the same landmark indices as before lm_eye_left = lm[36: 42] # left-clockwise lm_eye_right = lm[42: 48] # left-clockwise lm_mouth_outer = lm[48: 60] # left-clockwise # Calculate auxiliary vectors. eye_left = np.mean(lm_eye_left, axis=0) eye_right = np.mean(lm_eye_right, axis=0) eye_avg = (eye_left + eye_right) * 0.5 eye_to_eye = eye_right - eye_left mouth_left = lm_mouth_outer[0] mouth_right = lm_mouth_outer[6] mouth_avg = (mouth_left + mouth_right) * 0.5 eye_to_mouth = mouth_avg - eye_avg # Choose oriented crop rectangle. 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 # read image if isinstance(filepath, str): img = PIL.Image.open(filepath) else: img = PIL.Image.fromarray(filepath) output_size = 256 transform_size = 256 enable_padding = True # Shrink. shrink = int(np.floor(qsize / output_size * 0.5)) if shrink > 1: rsize = (int(np.rint(float(img.size[0]) / shrink)), int(np.rint(float(img.size[1]) / shrink))) img = img.resize(rsize, PIL.Image.ANTIALIAS) quad /= shrink qsize /= shrink # Crop. 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[0:2] # Pad. 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 = PIL.Image.fromarray(np.uint8(np.clip(np.rint(img), 0, 255)), 'RGB') quad += pad[:2] # Transform. img = img.transform((transform_size, transform_size), PIL.Image.QUAD, (quad + 0.5).flatten(), PIL.Image.BILINEAR) if output_size < transform_size: img = img.resize((output_size, output_size), PIL.Image.ANTIALIAS) return img def chunks(lst, n): """Yield successive n-sized chunks from lst.""" for i in range(0, len(lst), n): yield lst[i:i + n] def extract_on_paths(file_paths, face_detector): pid = mp.current_process().name print('\t{} is starting to extract on #{} images'.format(pid, len(file_paths))) tot_count = len(file_paths) count = 0 for file_path, res_path in file_paths: count += 1 if count % 100 == 0: print('{} done with {}/{}'.format(pid, count, tot_count)) try: res = align_face(file_path, face_detector) res = res.convert('RGB') os.makedirs(os.path.dirname(res_path), exist_ok=True) res.save(res_path) except Exception: continue print('\tDone!') def parse_args(): parser = ArgumentParser(add_help=False) parser.add_argument('--num_threads', type=int, default=1) parser.add_argument('--root_path', type=str, default='') args = parser.parse_args() return args def run(args): root_path = args.root_path out_crops_path = root_path + '_crops' if not os.path.exists(out_crops_path): os.makedirs(out_crops_path, exist_ok=True) file_paths = [] for root, dirs, files in os.walk(root_path): for file in files: file_path = os.path.join(root, file) fname = os.path.join(out_crops_path, os.path.relpath(file_path, root_path)) res_path = '{}.jpg'.format(os.path.splitext(fname)[0]) if os.path.splitext(file_path)[1] == '.txt' or os.path.exists(res_path): continue file_paths.append((file_path, res_path)) file_chunks = list(chunks(file_paths, int(math.ceil(len(file_paths) / args.num_threads)))) print(len(file_chunks)) pool = mp.Pool(args.num_threads) print('Running on {} paths\nHere we goooo'.format(len(file_paths))) tic = time.time() pool.starmap(extract_on_paths, [(chunk, face_detector) for chunk in file_chunks]) toc = time.time() print('Mischief managed in {}s'.format(toc - tic)) if __name__ == '__main__': # Initialize InsightFace face_detector = insightface.app.FaceAnalysis() face_detector.prepare(ctx_id=-1, det_size=(640, 640)) # ctx_id=-1 for CPU args = parse_args() run(args)