james-yip / attentionimageclass Goto Github PK

In this paper,I find that it has caffe version code,where could I find it?furtherMore,how about the result?does your result is better than the paper?Thanks

I try to implement ST + LSTM in TF and 6 LSTMs have been instantiated. Obviously it's wrong.
How to implement ST + LSTM as time-squence?

in code,
DIR_TRAIN_IMAGES = '../dataset/train2017/'
DIR_TEST_IMAGES = '../dataset/val2017/'
PATH_TRAIN_ANNFILE = '../dataset/annotations/instances_train2017.json'
PATH_TEST_ANNFILE = '../dataset/annotations/instances_val2017.json'
NUM_CATEGORIES = 80

Excuse me, is the data set in the code Microsoft COCO2014 data set?

I try to implement ST + LSTM in TF and 6 LSTMs have been instantiated. Obviously it's wrong.
How to implement ST + LSTM as time-squence in TF?

The threshold is 0.5 in the paper and 0.1 in the code. Why?

Hi, the LSTM used in the project is different from the artical which have an input modulate gate but the vanilla LSTM has not that gate. Dose the experiment results are the same as declared in the paper?

Hi there,

Thanks you for your reproduction of the paper.

But when I ran this code, I got 'No such file or directory: coco-label.txt'

Did you forget to commit this file?

Traceback (most recent call last):
File "D:\anaconda\lib\site-packages\urllib3\connection.py", line 157, in _new_conn
(self._dns_host, self.port), self.timeout, **extra_kw
File "D:\anaconda\lib\site-packages\urllib3\util\connection.py", line 84, in create_connection
raise err
File "D:\anaconda\lib\site-packages\urllib3\util\connection.py", line 74, in create_connection
sock.connect(sa)
ConnectionRefusedError: [WinError 10061] 由于目标计算机积极拒绝，无法连接。

M and scores are tensor. In pytorch, only parameter is trainable. So are M and scores not trainable?

PATH_MODEL_PARAMS = './params/params_no_priori.pkl'
where could I find this file?