diff --git a/README.md b/README.md index 197c443..b373a4b 100644 --- a/README.md +++ b/README.md @@ -79,15 +79,19 @@ export PYTHONPATH=$PYTHONPATH:{your local DFace root path} #### 训练mtcnn模型 -MTCNN主要有三个网络，叫做**PNet**, **RNet** 和 **ONet**。因此我们的训练过程也需要分三步先后进行。为了更好的实现效果，当前被训练的网络都将依赖于上一个训练好的网络来生成数据。所有的人脸数据集都来自　**[WIDER FACE](http://mmlab.ie.cuhk.edu.hk/projects/WIDERFace/)** 和 **[CelebA](http://mmlab.ie.cuhk.edu.hk/projects/CelebA.html)**。WIDER FACE仅提供了大量的人脸边框定位数据，而CelebA包含了人脸关键点定位数据。以下训练除了　生成ONet的人脸关键点训练数据和标注文件　该步骤使用CelebA数据集，其他一律使用WIDER FACE。如果使用wider face的 wider_face_train.mat 注解文件需要转换成txt格式的，我这里用h5py写了个 [转换脚本](https://gitee.com/kuaikuaikim/dface/blob/master/src/prepare_data/widerface_annotation_gen/transform.py) +MTCNN主要有三个网络，叫做**PNet**, **RNet** 和 **ONet**。因此我们的训练过程也需要分三步先后进行。为了更好的实现效果，当前被训练的网络都将依赖于上一个训练好的网络来生成数据。所有的人脸数据集都来自　**[WIDER FACE](http://mmlab.ie.cuhk.edu.hk/projects/WIDERFace/)** 和 **[CelebA](http://mmlab.ie.cuhk.edu.hk/projects/CelebA.html)**。WIDER FACE仅提供了大量的人脸边框定位数据，而CelebA包含了人脸关键点定位数据。以下训练除了　生成ONet的人脸关键点训练数据和标注文件　该步骤使用CelebA数据集，其他一律使用WIDER FACE。如果使用wider face的 wider_face_train.mat 注解文件需要转换成txt格式的，我这里用h5py写了个 [转换脚本](https://gitee.com/kuaikuaikim/dface/blob/master/dface/prepare_data/widerface_annotation_gen/transform.py). 这里我提供一个已经转换好的wider face注解文件 [anno_store/wider_origin_anno.txt](https://gitee.com/kuaikuaikim/dface/blob/master/anno_store/wider_origin_anno.txt), 以下训练过程参数名--anno_file默认就是使用该转换好的注解文件。 +* 创建 dface 训练数据临时目录，对应于以下所有的参数名　--dface_traindata_store +```shell +mkdir {your dface traindata folder} +``` * 生成PNet训练数据和标注文件 ```shell -python dface/prepare_data/gen_Pnet_train_data.py --dataset_path {your dataset path} --anno_file {your dataset original annotation path} +python dface/prepare_data/gen_Pnet_train_data.py --prefix_path {注解文件中图片的目录前缀} --dface_traindata_store {之前创建的dface训练数据临时目录}　--anno_file {wider face 注解文件,可以不填，默认使用anno_store/wider_origin_anno.txt} ``` * 乱序合并标注文件 @@ -104,7 +108,7 @@ python dface/train_net/train_p_net.py * 生成ＲNet训练数据和标注文件 ```shell -python dface/prepare_data/gen_Rnet_train_data.py --dataset_path {your dataset path} --anno_file {your dataset original annotation path} --pmodel_file {yout PNet model file trained before} +python dface/prepare_data/gen_Rnet_train_data.py --dface_traindata_store {之前创建的dface训练数据临时目录} --anno_file {wider face 注解文件,可以不填，默认使用anno_store/wider_origin_anno.txt} --pmodel_file {之前训练的Pnet模型文件} ``` * 乱序合并标注文件 @@ -121,16 +125,16 @@ python dface/train_net/train_r_net.py * 生成ONet训练数据和标注文件 ```shell -python dface/prepare_data/gen_Onet_train_data.py --dataset_path {your dataset path} --anno_file {your dataset original annotation path} --pmodel_file {yout PNet model file trained before} --rmodel_file {yout RNet model file trained before} +python dface/prepare_data/gen_Onet_train_data.py --dface_traindata_store {之前创建的dface训练数据临时目录} --anno_file {wider face 注解文件,可以不填，默认使用anno_store/wider_origin_anno.txt} --pmodel_file {之前训练的Pnet模型文件} --rmodel_file {之前训练的Rnet模型文件} ``` -* 生成ONet的人脸关键点训练数据和标注文件 +* 生成ONet的人脸五官关键点训练数据和标注文件 ```shell python dface/prepare_data/gen_landmark_48.py ``` -* 乱序合并标注文件(包括人脸关键点) +* 乱序合并标注文件(包括人脸五官关键点) ```shell python dface/prepare_data/assemble_onet_imglist.py diff --git a/dface/core/detect.py b/dface/core/detect.py index 5975f82..cdb53ac 100644 --- a/dface/core/detect.py +++ b/dface/core/detect.py @@ -623,7 +623,7 @@ face candidates:%d, current batch_size:%d"%(num_boxes, batch_size) t3 = time.time() - t t = time.time() - print "time cost " + '{:.3f}'.format(t1+t2+t3) + ' pnet {:.3f} rnet {:.3f} onet {:.3f}'.format(t1, t2, t3) + print("time cost " + '{:.3f}'.format(t1+t2+t3) + ' pnet {:.3f} rnet {:.3f} onet {:.3f}'.format(t1, t2, t3)) return boxes_align, landmark_align diff --git a/dface/core/imagedb.py b/dface/core/imagedb.py index be14ac2..839f0cf 100644 --- a/dface/core/imagedb.py +++ b/dface/core/imagedb.py @@ -137,7 +137,7 @@ class ImageDB(object): imdb: dict image database with flipped image annotations added """ - print 'append flipped images to imdb', len(imdb) + print('append flipped images to imdb', len(imdb)) for i in range(len(imdb)): imdb_ = imdb[i] m_bbox = imdb_['bbox_target'].copy() diff --git a/dface/prepare_data/assemble_onet_imglist.py b/dface/prepare_data/assemble_onet_imglist.py index 8e7b637..a9be5b9 100644 --- a/dface/prepare_data/assemble_onet_imglist.py +++ b/dface/prepare_data/assemble_onet_imglist.py @@ -22,4 +22,4 @@ if __name__ == '__main__': imglist_file = os.path.join(anno_dir, imglist_filename) chose_count = assemble.assemble_data(imglist_file ,anno_list) - print "PNet train annotation result file path:%s" % imglist_file + print("PNet train annotation result file path:%s" % imglist_file) diff --git a/dface/prepare_data/assemble_pnet_imglist.py b/dface/prepare_data/assemble_pnet_imglist.py index 5db7cc0..c93d36d 100644 --- a/dface/prepare_data/assemble_pnet_imglist.py +++ b/dface/prepare_data/assemble_pnet_imglist.py @@ -22,4 +22,4 @@ if __name__ == '__main__': imglist_file = os.path.join(anno_dir, imglist_filename) chose_count = assemble.assemble_data(imglist_file ,anno_list) - print "PNet train annotation result file path:%s" % imglist_file + print("PNet train annotation result file path:%s" % imglist_file) diff --git a/dface/prepare_data/assemble_rnet_imglist.py b/dface/prepare_data/assemble_rnet_imglist.py index 2e6ba35..417547e 100644 --- a/dface/prepare_data/assemble_rnet_imglist.py +++ b/dface/prepare_data/assemble_rnet_imglist.py @@ -22,4 +22,4 @@ if __name__ == '__main__': imglist_file = os.path.join(anno_dir, imglist_filename) chose_count = assemble.assemble_data(imglist_file ,anno_list) - print "PNet train annotation result file path:%s" % imglist_file + print("PNet train annotation result file path:%s" % imglist_file) diff --git a/dface/prepare_data/gen_Onet_train_data.py b/dface/prepare_data/gen_Onet_train_data.py index f8dbfa6..8cb1c1c 100644 --- a/dface/prepare_data/gen_Onet_train_data.py +++ b/dface/prepare_data/gen_Onet_train_data.py @@ -9,7 +9,7 @@ import time import os import cPickle from dface.core.utils import convert_to_square,IoU -import dface.config +import dface.config as config import dface.core.vision as vision def gen_onet_data(data_dir, anno_file, pnet_model_file, rnet_model_file, prefix_path='', use_cuda=True, vis=False): @@ -27,7 +27,7 @@ def gen_onet_data(data_dir, anno_file, pnet_model_file, rnet_model_file, prefix_ for databatch in image_reader: if batch_idx % 100 == 0: - print "%d images done" % batch_idx + print("%d images done" % batch_idx) im = databatch t = time.time() @@ -89,7 +89,7 @@ def gen_onet_sample_data(data_dir,anno_file,det_boxs_file): im_idx_list = list() gt_boxes_list = list() num_of_images = len(annotations) - print "processing %d images in total" % num_of_images + print("processing %d images in total" % num_of_images) for annotation in annotations: annotation = annotation.strip().split(' ') @@ -112,7 +112,7 @@ def gen_onet_sample_data(data_dir,anno_file,det_boxs_file): det_handle = open(det_boxs_file, 'r') det_boxes = cPickle.load(det_handle) - print len(det_boxes), num_of_images + print(len(det_boxes), num_of_images) assert len(det_boxes) == num_of_images, "incorrect detections or ground truths" # index of neg, pos and part face, used as their image names @@ -122,7 +122,7 @@ def gen_onet_sample_data(data_dir,anno_file,det_boxs_file): image_done = 0 for im_idx, dets, gts in zip(im_idx_list, det_boxes, gt_boxes_list): if image_done % 100 == 0: - print "%d images done" % image_done + print("%d images done" % image_done) image_done += 1 if dets.shape[0] == 0: @@ -194,17 +194,17 @@ def parse_args(): parser = argparse.ArgumentParser(description='Test mtcnn', formatter_class=argparse.ArgumentDefaultsHelpFormatter) - parser.add_argument('--dataset_path', dest='dataset_path', help='dataset folder', + parser.add_argument('--dface_traindata_store', dest='traindata_store', help='dface train data temporary folder,include 12,24,48/postive,negative,part,landmark', default='../data/wider/', type=str) - parser.add_argument('--anno_file', dest='annotation_file', help='output data folder', - default='../data/wider/anno.txt', type=str) + parser.add_argument('--anno_file', dest='annotation_file', help='wider face original annotation file', + default=os.path.join(config.ANNO_STORE_DIR,"wider_origin_anno.txt"), type=str) parser.add_argument('--pmodel_file', dest='pnet_model_file', help='PNet model file path', - default='/idata/workspace/mtcnn/model_store/pnet_epoch_5best.pt', type=str) + default='/idata/workspace/dface/model_store/pnet_epoch.pt', type=str) parser.add_argument('--rmodel_file', dest='rnet_model_file', help='RNet model file path', - default='/idata/workspace/mtcnn/model_store/rnet_epoch_1.pt', type=str) + default='/idata/workspace/dface/model_store/rnet_epoch.pt', type=str) parser.add_argument('--gpu', dest='use_cuda', help='with gpu', default=config.USE_CUDA, type=bool) - parser.add_argument('--prefix_path', dest='prefix_path', help='image prefix root path', + parser.add_argument('--prefix_path', dest='prefix_path', help='annotation file image prefix root path', default='', type=str) args = parser.parse_args() @@ -214,7 +214,7 @@ def parse_args(): if __name__ == '__main__': args = parse_args() - gen_onet_data(args.dataset_path, args.annotation_file, args.pnet_model_file, args.rnet_model_file, args.prefix_path, args.use_cuda) + gen_onet_data(args.traindata_store, args.annotation_file, args.pnet_model_file, args.rnet_model_file, args.prefix_path, args.use_cuda) diff --git a/dface/prepare_data/gen_Pnet_train_data.py b/dface/prepare_data/gen_Pnet_train_data.py index c297fe9..1d2b598 100644 --- a/dface/prepare_data/gen_Pnet_train_data.py +++ b/dface/prepare_data/gen_Pnet_train_data.py @@ -6,7 +6,7 @@ import numpy.random as npr from dface.core.utils import IoU import dface.config as config -def gen_pnet_data(data_dir,anno_file): +def gen_pnet_data(data_dir,anno_file,prefix): neg_save_dir = os.path.join(data_dir,"12/negative") pos_save_dir = os.path.join(data_dir,"12/positive") @@ -32,7 +32,7 @@ def gen_pnet_data(data_dir,anno_file): annotations = f.readlines() num = len(annotations) - print "%d pics in total" % num + print("%d pics in total" % num) p_idx = 0 n_idx = 0 d_idx = 0 @@ -40,13 +40,13 @@ def gen_pnet_data(data_dir,anno_file): box_idx = 0 for annotation in annotations: annotation = annotation.strip().split(' ') - im_path = annotation[0] + im_path = os.path.join(prefix,annotation[0]) bbox = map(float, annotation[1:]) boxes = np.array(bbox, dtype=np.int32).reshape(-1, 4) img = cv2.imread(im_path) idx += 1 if idx % 100 == 0: - print idx, "images done" + print(idx, "images done") height, width, channel = img.shape @@ -145,7 +145,7 @@ def gen_pnet_data(data_dir,anno_file): cv2.imwrite(save_file, resized_im) d_idx += 1 box_idx += 1 - print "%s images done, pos: %s part: %s neg: %s"%(idx, p_idx, d_idx, n_idx) + print("%s images done, pos: %s part: %s neg: %s"%(idx, p_idx, d_idx, n_idx)) f1.close() f2.close() @@ -157,16 +157,19 @@ def parse_args(): parser = argparse.ArgumentParser(description='Test mtcnn', formatter_class=argparse.ArgumentDefaultsHelpFormatter) - parser.add_argument('--dataset_path', dest='dataset_path', help='dataset folder', + parser.add_argument('--dface_traindata_store', dest='traindata_store', help='dface train data temporary folder,include 12,24,48/postive,negative,part,landmark', default='../data/wider/', type=str) - parser.add_argument('--anno_file', dest='annotation_file', help='dataset original annotation file', - default='../data/wider/anno.txt', type=str) - parser.add_argument('--prefix_path', dest='prefix_path', help='image prefix root path', + parser.add_argument('--anno_file', dest='annotation_file', help='wider face original annotation file', + default=os.path.join(config.ANNO_STORE_DIR,"wider_origin_anno.txt"), type=str) + parser.add_argument('--prefix_path', dest='prefix_path', help='annotation file image prefix root path', default='', type=str) + + + args = parser.parse_args() return args if __name__ == '__main__': args = parse_args() - gen_pnet_data(args.dataset_path,args.annotation_file) \ No newline at end of file + gen_pnet_data(args.traindata_store,args.annotation_file,args.prefix_path) \ No newline at end of file diff --git a/dface/prepare_data/gen_Rnet_train_data.py b/dface/prepare_data/gen_Rnet_train_data.py index fd6ff66..4f8888f 100644 --- a/dface/prepare_data/gen_Rnet_train_data.py +++ b/dface/prepare_data/gen_Rnet_train_data.py @@ -28,7 +28,7 @@ def gen_rnet_data(data_dir, anno_file, pnet_model_file, prefix_path='', use_cuda for databatch in image_reader: if batch_idx % 100 == 0: - print "%d images done" % batch_idx + print ("%d images done" % batch_idx) im = databatch t = time.time() @@ -58,11 +58,11 @@ def gen_rnet_data(data_dir, anno_file, pnet_model_file, prefix_path='', use_cuda cPickle.dump(all_boxes, f, cPickle.HIGHEST_PROTOCOL) - gen_rnet_sample_data(data_dir,anno_file,save_file) + gen_rnet_sample_data(data_dir,anno_file,save_file,prefix_path) -def gen_rnet_sample_data(data_dir,anno_file,det_boxs_file): +def gen_rnet_sample_data(data_dir,anno_file,det_boxs_file,prefix_path): neg_save_dir = os.path.join(data_dir, "24/negative") pos_save_dir = os.path.join(data_dir, "24/positive") @@ -85,11 +85,11 @@ def gen_rnet_sample_data(data_dir,anno_file,det_boxs_file): im_idx_list = list() gt_boxes_list = list() num_of_images = len(annotations) - print "processing %d images in total" % num_of_images + print ("processing %d images in total" % num_of_images) for annotation in annotations: annotation = annotation.strip().split(' ') - im_idx = annotation[0] + im_idx = os.path.join(prefix_path,annotation[0]) boxes = map(float, annotation[1:]) boxes = np.array(boxes, dtype=np.float32).reshape(-1, 4) @@ -108,7 +108,7 @@ def gen_rnet_sample_data(data_dir,anno_file,det_boxs_file): det_handle = open(det_boxs_file, 'r') det_boxes = cPickle.load(det_handle) - print len(det_boxes), num_of_images + print(len(det_boxes), num_of_images) assert len(det_boxes) == num_of_images, "incorrect detections or ground truths" # index of neg, pos and part face, used as their image names @@ -118,7 +118,7 @@ def gen_rnet_sample_data(data_dir,anno_file,det_boxs_file): image_done = 0 for im_idx, dets, gts in zip(im_idx_list, det_boxes, gt_boxes_list): if image_done % 100 == 0: - print "%d images done" % image_done + print("%d images done" % image_done) image_done += 1 if dets.shape[0] == 0: @@ -194,15 +194,15 @@ def parse_args(): parser = argparse.ArgumentParser(description='Test mtcnn', formatter_class=argparse.ArgumentDefaultsHelpFormatter) - parser.add_argument('--dataset_path', dest='dataset_path', help='dataset folder', + parser.add_argument('--dface_traindata_store', dest='traindata_store', help='dface train data temporary folder,include 12,24,48/postive,negative,part,landmark', default='../data/wider/', type=str) - parser.add_argument('--anno_file', dest='annotation_file', help='dataset original annotation file', - default='../data/wider/anno.txt', type=str) + parser.add_argument('--anno_file', dest='annotation_file', help='wider face original annotation file', + default=os.path.join(config.ANNO_STORE_DIR,"wider_origin_anno.txt"), type=str) parser.add_argument('--pmodel_file', dest='pnet_model_file', help='PNet model file path', - default='/idata/workspace/mtcnn/model_store/pnet_epoch_5best.pt', type=str) + default='/idata/workspace/dface/model_store/pnet_epoch.pt', type=str) parser.add_argument('--gpu', dest='use_cuda', help='with gpu', default=config.USE_CUDA, type=bool) - parser.add_argument('--prefix_path', dest='prefix_path', help='image prefix root path', + parser.add_argument('--prefix_path', dest='prefix_path', help='annotation file image prefix root path', default='', type=str) @@ -213,7 +213,7 @@ def parse_args(): if __name__ == '__main__': args = parse_args() - gen_rnet_data(args.dataset_path, args.annotation_file, args.pnet_model_file, args.prefix_path, args.use_cuda) + gen_rnet_data(args.traindata_store, args.annotation_file, args.pnet_model_file, args.prefix_path, args.use_cuda) diff --git a/dface/prepare_data/gen_landmark_12.py b/dface/prepare_data/gen_landmark_12.py index 2c4a34f..9424ff3 100644 --- a/dface/prepare_data/gen_landmark_12.py +++ b/dface/prepare_data/gen_landmark_12.py @@ -34,7 +34,7 @@ def gen_data(anno_file, data_dir, prefix): annotations = f2.readlines() num = len(annotations) - print "%d pics in total" % num + print("%d pics in total" % num) l_idx =0 idx = 0 @@ -68,7 +68,7 @@ def gen_data(anno_file, data_dir, prefix): idx = idx + 1 if idx % 100 == 0: - print "%d images done, landmark images: %d"%(idx,l_idx) + print("%d images done, landmark images: %d"%(idx,l_idx)) x1, y1, x2, y2 = gt_box @@ -137,11 +137,11 @@ def parse_args(): parser = argparse.ArgumentParser(description='Test mtcnn', formatter_class=argparse.ArgumentDefaultsHelpFormatter) - parser.add_argument('--dataset_path', dest='dataset_path', help='dataset folder', + parser.add_argument('--dface_traindata_store', dest='traindata_store', help='dataset folder', default='../data/wider/', type=str) parser.add_argument('--anno_file', dest='annotation_file', help='dataset original annotation file', default='../data/wider/anno.txt', type=str) - parser.add_argument('--prefix_path', dest='prefix_path', help='image prefix root path', + parser.add_argument('--prefix_path', dest='prefix_path', help='annotation file image prefix root path', default='../data/', type=str) @@ -151,6 +151,6 @@ def parse_args(): if __name__ == '__main__': args = parse_args() - gen_data(args.annotation_file, args.dataset_path, args.prefix_path) + gen_data(args.annotation_file, args.traindata_store, args.prefix_path) diff --git a/dface/prepare_data/gen_landmark_24.py b/dface/prepare_data/gen_landmark_24.py index 590be8f..1b2b4b1 100644 --- a/dface/prepare_data/gen_landmark_24.py +++ b/dface/prepare_data/gen_landmark_24.py @@ -35,7 +35,7 @@ def gen_data(anno_file, data_dir, prefix): annotations = f2.readlines() num = len(annotations) - print "%d total images" % num + print("%d total images" % num) l_idx =0 idx = 0 @@ -67,7 +67,7 @@ def gen_data(anno_file, data_dir, prefix): idx = idx + 1 if idx % 100 == 0: - print "%d images done, landmark images: %d"%(idx,l_idx) + print("%d images done, landmark images: %d"%(idx,l_idx)) x1, y1, x2, y2 = gt_box @@ -135,11 +135,11 @@ def parse_args(): parser = argparse.ArgumentParser(description='Test mtcnn', formatter_class=argparse.ArgumentDefaultsHelpFormatter) - parser.add_argument('--dataset_path', dest='dataset_path', help='dataset folder', + parser.add_argument('--dface_traindata_store', dest='traindata_store', help='dataset folder', default='/idata/data/wider/', type=str) parser.add_argument('--anno_file', dest='annotation_file', help='dataset original annotation file', default='/idata/data/trainImageList.txt', type=str) - parser.add_argument('--prefix_path', dest='prefix_path', help='image prefix root path', + parser.add_argument('--prefix_path', dest='prefix_path', help='annotation file image prefix root path', default='/idata/data', type=str) @@ -149,6 +149,6 @@ def parse_args(): if __name__ == '__main__': args = parse_args() - gen_data(args.annotation_file, args.dataset_path, args.prefix_path) + gen_data(args.annotation_file, args.traindata_store, args.prefix_path) diff --git a/dface/prepare_data/gen_landmark_48.py b/dface/prepare_data/gen_landmark_48.py index 826d82c..01571f5 100644 --- a/dface/prepare_data/gen_landmark_48.py +++ b/dface/prepare_data/gen_landmark_48.py @@ -34,7 +34,7 @@ def gen_data(anno_file, data_dir, prefix): annotations = f2.readlines() num = len(annotations) - print "%d total images" % num + print("%d total images" % num) l_idx =0 idx = 0 @@ -66,7 +66,7 @@ def gen_data(anno_file, data_dir, prefix): idx = idx + 1 if idx % 100 == 0: - print "%d images done, landmark images: %d"%(idx,l_idx) + print("%d images done, landmark images: %d"%(idx,l_idx)) x1, y1, x2, y2 = gt_box @@ -134,11 +134,11 @@ def parse_args(): parser = argparse.ArgumentParser(description='Test mtcnn', formatter_class=argparse.ArgumentDefaultsHelpFormatter) - parser.add_argument('--dataset_path', dest='dataset_path', help='dataset folder', + parser.add_argument('--dface_traindata_store', dest='traindata_store', help='dataset folder', default='/idata/data/wider/', type=str) parser.add_argument('--anno_file', dest='annotation_file', help='dataset original annotation file', default='/idata/data/trainImageList.txt', type=str) - parser.add_argument('--prefix_path', dest='prefix_path', help='image prefix root path', + parser.add_argument('--prefix_path', dest='prefix_path', help='annotation file image prefix root path', default='/idata/data', type=str) @@ -148,6 +148,6 @@ def parse_args(): if __name__ == '__main__': args = parse_args() - gen_data(args.annotation_file, args.dataset_path, args.prefix_path) + gen_data(args.annotation_file, args.traindata_store, args.prefix_path) diff --git a/dface/prepare_data/gen_landmark_net_48.py b/dface/prepare_data/gen_landmark_net_48.py index 20807f4..87fd097 100644 --- a/dface/prepare_data/gen_landmark_net_48.py +++ b/dface/prepare_data/gen_landmark_net_48.py @@ -27,7 +27,7 @@ def gen_landmark48_data(data_dir, anno_file, pnet_model_file, rnet_model_file, p for databatch in image_reader: if batch_idx % 100 == 0: - print "%d images done" % batch_idx + print("%d images done" % batch_idx) im = databatch @@ -91,7 +91,7 @@ def gen_sample_data(data_dir, anno_file, det_boxs_file, prefix_path =''): gt_boxes_list = list() gt_landmark_list = list() num_of_images = len(annotations) - print "processing %d images in total" % num_of_images + print("processing %d images in total" % num_of_images) for annotation in annotations: annotation = annotation.strip().split(' ') @@ -117,7 +117,7 @@ def gen_sample_data(data_dir, anno_file, det_boxs_file, prefix_path =''): det_handle = open(det_boxs_file, 'r') det_boxes = cPickle.load(det_handle) - print len(det_boxes), num_of_images + print(len(det_boxes), num_of_images) assert len(det_boxes) == num_of_images, "incorrect detections or ground truths" # index of neg, pos and part face, used as their image names @@ -125,7 +125,7 @@ def gen_sample_data(data_dir, anno_file, det_boxs_file, prefix_path =''): image_done = 0 for im_idx, dets, gts, landmark in zip(im_idx_list, det_boxes, gt_boxes_list, gt_landmark_list): if image_done % 100 == 0: - print "%d images done" % image_done + print("%d images done" % image_done) image_done += 1 if dets.shape[0] == 0: @@ -208,7 +208,7 @@ def parse_args(): parser = argparse.ArgumentParser(description='Test mtcnn', formatter_class=argparse.ArgumentDefaultsHelpFormatter) - parser.add_argument('--dataset_path', dest='dataset_path', help='dataset folder', + parser.add_argument('--dface_traindata_store', dest='traindata_store', help='dataset folder', default='../data/wider/', type=str) parser.add_argument('--anno_file', dest='annotation_file', help='output data folder', default='../data/wider/anno.txt', type=str) @@ -228,7 +228,7 @@ def parse_args(): if __name__ == '__main__': args = parse_args() - gen_landmark48_data(args.dataset_path, args.annotation_file, args.pnet_model_file, args.rnet_model_file, args.prefix_path, args.use_cuda) + gen_landmark48_data(args.traindata_store, args.annotation_file, args.pnet_model_file, args.rnet_model_file, args.prefix_path, args.use_cuda) diff --git a/dface/prepare_data/widerface_annotation_gen/__init__.py b/dface/prepare_data/widerface_annotation_gen/__init__.py new file mode 100644 index 0000000..e69de29 diff --git a/dface/prepare_data/widerface_annotation_gen/transform.py b/dface/prepare_data/widerface_annotation_gen/transform.py index ec923df..847794a 100644 --- a/dface/prepare_data/widerface_annotation_gen/transform.py +++ b/dface/prepare_data/widerface_annotation_gen/transform.py @@ -1,4 +1,4 @@ -from wider_loader import WIDER +from dface.prepare_data.widerface_annotation_gen.wider_loader import WIDER import cv2 import time @@ -17,7 +17,7 @@ wider = WIDER(file_to_label, path_to_image) line_count = 0 box_count = 0 -print 'start transforming....' +print('start transforming....') t = time.time() with open(target_file, 'w+') as f: @@ -37,10 +37,10 @@ with open(target_file, 'w+') as f: f.write(line_str) st = time.time()-t -print 'end transforming' +print('end transforming') -print 'spend time:%ld'%st -print 'total line(images):%d'%line_count -print 'total boxes(faces):%d'%box_count +print('spend time:%ld'%st) +print('total line(images):%d'%line_count) +print('total boxes(faces):%d'%box_count) diff --git a/dface/train_net/train.py b/dface/train_net/train.py index f19f69f..129edd2 100644 --- a/dface/train_net/train.py +++ b/dface/train_net/train.py @@ -84,7 +84,7 @@ def train_pnet(model_store_path, end_epoch,imdb, show3 = box_offset_loss.data.tolist()[0] show5 = all_loss.data.tolist()[0] - print "%s : Epoch: %d, Step: %d, accuracy: %s, det loss: %s, bbox loss: %s, all_loss: %s, lr:%s "%(datetime.datetime.now(),cur_epoch,batch_idx, show1,show2,show3,show5,base_lr) + print("%s : Epoch: %d, Step: %d, accuracy: %s, det loss: %s, bbox loss: %s, all_loss: %s, lr:%s "%(datetime.datetime.now(),cur_epoch,batch_idx, show1,show2,show3,show5,base_lr)) accuracy_list.append(accuracy) cls_loss_list.append(cls_loss) bbox_loss_list.append(box_offset_loss) @@ -103,7 +103,7 @@ def train_pnet(model_store_path, end_epoch,imdb, show7 = cls_loss_avg.data.tolist()[0] show8 = bbox_loss_avg.data.tolist()[0] - print "Epoch: %d, accuracy: %s, cls loss: %s, bbox loss: %s" % (cur_epoch, show6, show7, show8) + print("Epoch: %d, accuracy: %s, cls loss: %s, bbox loss: %s" % (cur_epoch, show6, show7, show8)) torch.save(net.state_dict(), os.path.join(model_store_path,"pnet_epoch_%d.pt" % cur_epoch)) torch.save(net, os.path.join(model_store_path,"pnet_epoch_model_%d.pkl" % cur_epoch)) @@ -169,7 +169,7 @@ def train_rnet(model_store_path, end_epoch,imdb, # show4 = landmark_loss.data.tolist()[0] show5 = all_loss.data.tolist()[0] - print "%s : Epoch: %d, Step: %d, accuracy: %s, det loss: %s, bbox loss: %s, all_loss: %s, lr:%s "%(datetime.datetime.now(), cur_epoch, batch_idx, show1, show2, show3, show5, base_lr) + print("%s : Epoch: %d, Step: %d, accuracy: %s, det loss: %s, bbox loss: %s, all_loss: %s, lr:%s "%(datetime.datetime.now(), cur_epoch, batch_idx, show1, show2, show3, show5, base_lr)) accuracy_list.append(accuracy) cls_loss_list.append(cls_loss) bbox_loss_list.append(box_offset_loss) @@ -190,7 +190,7 @@ def train_rnet(model_store_path, end_epoch,imdb, show8 = bbox_loss_avg.data.tolist()[0] # show9 = landmark_loss_avg.data.tolist()[0] - print "Epoch: %d, accuracy: %s, cls loss: %s, bbox loss: %s" % (cur_epoch, show6, show7, show8) + print("Epoch: %d, accuracy: %s, cls loss: %s, bbox loss: %s" % (cur_epoch, show6, show7, show8)) torch.save(net.state_dict(), os.path.join(model_store_path,"rnet_epoch_%d.pt" % cur_epoch)) torch.save(net, os.path.join(model_store_path,"rnet_epoch_model_%d.pkl" % cur_epoch)) @@ -275,7 +275,7 @@ def train_onet(model_store_path, end_epoch,imdb, show8 = bbox_loss_avg.data.tolist()[0] show9 = landmark_loss_avg.data.tolist()[0] - print "Epoch: %d, accuracy: %s, cls loss: %s, bbox loss: %s, landmark loss: %s " % (cur_epoch, show6, show7, show8, show9) + print("Epoch: %d, accuracy: %s, cls loss: %s, bbox loss: %s, landmark loss: %s " % (cur_epoch, show6, show7, show8, show9)) torch.save(net.state_dict(), os.path.join(model_store_path,"onet_epoch_%d.pt" % cur_epoch)) torch.save(net, os.path.join(model_store_path,"onet_epoch_model_%d.pkl" % cur_epoch)) diff --git a/dface/train_net/train_o_net.py b/dface/train_net/train_o_net.py index 49ca261..703ab5a 100644 --- a/dface/train_net/train_o_net.py +++ b/dface/train_net/train_o_net.py @@ -42,7 +42,7 @@ def parse_args(): if __name__ == '__main__': args = parse_args() - print 'train ONet argument:' + print('train ONet argument:') print args diff --git a/dface/train_net/train_p_net.py b/dface/train_net/train_p_net.py index 79aa16f..53df578 100644 --- a/dface/train_net/train_p_net.py +++ b/dface/train_net/train_p_net.py @@ -2,7 +2,7 @@ import argparse import sys from dface.core.imagedb import ImageDB from train import train_pnet -import dface.config +import dface.config as config import os @@ -42,7 +42,7 @@ def parse_args(): if __name__ == '__main__': args = parse_args() - print 'train Pnet argument:' + print('train Pnet argument:') print args train_net(annotation_file=args.annotation_file, model_store_path=args.model_store_path, diff --git a/dface/train_net/train_r_net.py b/dface/train_net/train_r_net.py index 607a970..ac3ad00 100644 --- a/dface/train_net/train_r_net.py +++ b/dface/train_net/train_r_net.py @@ -2,7 +2,7 @@ import argparse import sys from dface.core.imagedb import ImageDB import train as train -import dface.config +import dface.config as config import os @@ -42,7 +42,7 @@ def parse_args(): if __name__ == '__main__': args = parse_args() - print 'train Rnet argument:' + print('train Rnet argument:') print args diff --git a/test_image.py b/test_image.py index 8eac11a..e4f8a65 100644 --- a/test_image.py +++ b/test_image.py @@ -3,7 +3,6 @@ from dface.core.detect import create_mtcnn_net, MtcnnDetector from dface.core import vision - if __name__ == '__main__': # refer to your local model path