search

sairin1202 / Commonsense-Knowledge-Aware-Concept-Selection-For-Diverse-and-Informative-Visual-Storytelling

The implement of Commonsense Knowledge Aware Concept Selection For Diverse and Informative Visual Storytelling
MIT License
11 stars 2 forks source link

The paper description differs from the provided code implementation？ #7

Open Thinking-more opened 2 years ago

Thinking-more commented 2 years ago

Hello, I noticed in your paper when I reproduced your results that you say "During training, we follow Jung et al. (2020) to blind one of the images starting from the 50-th epoch and increase the blinding into two images from epoch 80. The training stops at epoch 100.", but I don't see this part of the code in the code you provided, is it possible to add it if possible? In your paper, you said "VIST dataset (Huang et al. 2016), which consists of 40101 samples for training 5050 test samples. and one corresponding gold story." but the results you provided only have 1010, why is that?

xinli2008 commented 2 years ago

I also find the second question~ In VIST dataset, one album_id correspond to 5 stories. But in the file album2story.json and graph2story.json, album_id correspond to 1 story! So the results only have 1010 stories rather than 5050 stories!

Thinking-more commented 2 years ago

I also find the second question~ In VIST dataset, one album_id correspond to 5 stories. But in the file album2story.json and graph2story.json, album_id correspond to 1 story! So the results only have 1010 stories rather than 5050 stories!

It seems that you also reserach the Visual Storytelling. If possible，maybe we can talk further.

xinli2008 commented 2 years ago

@RENJZ1026 It is my pleasure

Thinking-more commented 2 years ago

@RENJZ1026 It is my pleasure

Good! rj2521715866@gmail.com, It's my email

xinli2008 commented 2 years ago

@RENJZ1026 It is my pleasure

Good! rj2521715866@gmail.com, It's my email

QQ？