Search results

How to process the multiple choice answer

Status: Open.

#35 In GT-Vision-Lab/VQA_LSTM_CNN;

· WangWenshan opened on Jan 10, 2019

Abstract scene parameters num_ans and num_output

Status: Open.

#34 In GT-Vision-Lab/VQA_LSTM_CNN;

· sanjass opened on Dec 7, 2018

UNk Token

Status: Open.

#33 In GT-Vision-Lab/VQA_LSTM_CNN;

· samarIbrahem opened on Feb 5, 2018

no clue

Status: Open.

#32 In GT-Vision-Lab/VQA_LSTM_CNN;

· jijibn opened on Jul 15, 2017

Number of pretrained image features not matching with number of images in COCO

Status: Open.

#29 In GT-Vision-Lab/VQA_LSTM_CNN;

· franroldans opened on May 30, 2017

Fail to repeat the accuracy of the pretrained VGG model

Status: Open.

#27 In GT-Vision-Lab/VQA_LSTM_CNN;

· xhzhao opened on Dec 8, 2016

would you tell me more about the parameter and dataset?

Status: Open.

#26 In GT-Vision-Lab/VQA_LSTM_CNN;

· SeekPoint opened on Oct 27, 2016

Providing feedback through correct answer

Status: Open.

#25 In GT-Vision-Lab/VQA_LSTM_CNN;

· goodrahstar opened on Oct 12, 2016

Might be remove the second term of output in LSTM

Status: Open.

#21 In GT-Vision-Lab/VQA_LSTM_CNN;

· haooooooqi opened on Apr 25, 2016

require 'cunn' and 'cutorch' in CPU mode

Status: Open.

#19 In GT-Vision-Lab/VQA_LSTM_CNN;

· NightFury13 opened on Apr 4, 2016

Number of training picture

Status: Open.

#17 In GT-Vision-Lab/VQA_LSTM_CNN;

· andrewliao11 opened on Mar 16, 2016

Ideas for NLP pre-processing and feature engineering

Status: Open.

#14 In GT-Vision-Lab/VQA_LSTM_CNN;

· honnibal opened on Mar 6, 2016