yanzi6969
yanzi6969
The example label for the 3 types data: positive face: face detection: 1; face landmark: [0.1,0.2,0.3,0.4,0.5,0.1,0.2,0.3,0.4,0.5]; face regression: [0.1,0.1,0.1,0.1]. part of the face: face detection: 1; face landmark: [0.1,0.2,0.3,0.4,0.5,0.1,0.2,0.3,0.4,0.5]; face...
landmark取的是哪些特征呢？能否麻烦您把 label_path = '../dataset/label.txt' landmark_path = '../dataset/landmark.txt' regression_box_path = '../dataset/regression_box.txt' crop_image_path = '../dataset/crop_image.txt' 这几个txt共享下可以吗？
@zfc929 请教下 ，如果我不训练landmark，只训练人脸框，该怎么设置标签呢？
@linsonwang 我理解是没做区分的，也可以训练出结果。
能共享下 label_path = '../dataset/label.txt' landmark_path = '../dataset/landmark.txt' regression_box_path = '../dataset/regression_box.txt' crop_image_path = '../dataset/crop_image.txt' 这几个txt吗？实在没搞懂具体格式该怎样，谢谢
@foreverYoungGitHub face regression: [0.1,0.1,0.1,0.1]. left_x,left_y,width,height or left_x, left_y, ritht_x, ritht_y?
use celeba_crop get the rectangle.txt , like this: 0.03 0.04 0.04 -0.04 0.02 0 0 0.05 -0.04 -0.05 0.02 0.02 0 0.03 0.01 0.02 -0.02 0.02 0.04 -0.03 0.02 0.02...
So slow... Cant use without gpu