search

Pay20Y / FOTS_TF

This an implementation of FOTS with tensorflow
GNU General Public License v3.0
181 stars 59 forks source link

Specify epochs? #57

Open daniiki opened 4 years ago

daniiki commented 4 years ago

Hi, where can you specify the number of epochs for training? Thanks!

Pay20Y commented 4 years ago

Sorry, the epochs can not be specified directlly in this code, please calculate the corresponding number of iterations.

daniiki commented 4 years ago

Thanks for the fast answer. Can you tell me how to compute the recognition accuracy / precision? I computed the detection accuracy like this:

d_correct_prediction = tf.equal(tf.argmax(f_geometry, 1), tf.argmax(input_geo_maps, 1)) detection_accuracy = tf.reduce_mean(tf.cast(d_correct_prediction, tf.float32))

but the calculation of the recognition accuracy seems to be more complicated, since _inputtranscription is a sparse tensor.

Pay20Y commented 4 years ago

Hi, you can modify the data generator and make it yield another string of label. I recommend that you refer to the ICDAR official evaluation script.