search

Mmmofan / YOLO_6D

Implement Singleshotpose(Yolo 6d) in TensorFlow
55 stars 9 forks source link

tensorflow.python.framework.errors_impl.InvalidArgumentError #1

Open huobaojiajie666 opened 5 years ago

huobaojiajie666 commented 5 years ago

Hi,thank for your code,but I got this error when I was training.How can I solve it to achieve training? tensorflow.python.framework.errors_impl.InvalidArgumentError: You must feed a value for placeholder tensor 'target' with dtype float and shape [4,21] [[{{node target}}]] [[{{node global_step/initial_value}}]]

ghoshaw commented 5 years ago

The line "cur_gt_corners = tf.transpose(tf.tile(tf.Variable([[gx0, gy0, gx1, gy1, gx2, gy2, gx3, gy3, gx4, gy4, gx5, gy5, gx6, gy6, gx7, gy7, gx8, gy8]], trainable=False), (nAnchors, 1)), (1, 0)) " in 316 in yolo_6d.py should change the tf.Variable to tf.stack, then the error disappeared. by the way, I think if your label is saved in (x0, y0, x1, y1) as singleshot does, then in 105 of linemod.py, center_x should equal to gt_label[0]nW, not gt_label[0]nH. so is the following lines. Hope that helps!