Kenneth-Wong

You may have a try: change the original "lstm_cell = tf.contrib.rnn.BasicLSTMCell(lstm_dim, state_is_tuple=True) outputs2_raw, _ = tf.nn.bidirectional_dynamic_rnn(lstm_cell, lstm_cell, outputs1, seq_length, dtype=tf.float32, time_major=True, scope="bidirectional_lstm2")" into “lstm_cell_2 = tf.contrib.rnn.BasicLSTMCell(lstm_dim, state_is_tuple=True) outputs2_raw, _ =...

Oh I see that in `preprocess.py`, you use the `relational_caption.json` as input. It seems that the json file is from the raw VG dataset. However, on the VG website, I...

Hello. I am not coming for this question. I just want to ask what's your MAP after training 160 epochs on PASCAL VOC 20017&2012 trainval set. My MAP is very...

You are right. I modified the codes and reshaped the groundtruth boxes according to the input image shapes. There were not any errors reported and the MAP are improved a...

I think this is because your source prompt and target prompt are almost the same, and the processed results of the tokenizer are so similar to the space-splited ones. For...