search

j-min / Adversarial_Video_Summary

Unofficial PyTorch Implementation of SUM-GAN from "Unsupervised Video Summarization with Adversarial LSTM Networks" (CVPR 2017)
239 stars 61 forks source link

poor results applying video summarization on BDD100 dataset #16

Open EmnamoR opened 4 years ago

EmnamoR commented 4 years ago
Screenshot 2019-10-23 at 14 10 35

I am trying to apply the network on BDD100 dataset which is for drives so c_loss is -Gan_loss

in the paper we should :

  1. For learning {θs, θe}, minimize (Lreconst+Lprior+Lsparsity). ==> s_e_epoch
  2. For learning θd, minimize (Lreconst+LGAN). d_epoch
  3. For learning θc, maximize LGAN. which is -c_loss so minimize c_epoch

but i am having this behaviour? what could be the problem ?

shineYuSong commented 4 years ago

have you realized the original project? can you transfer the dataset to me

kundezui commented 3 years ago

Could you tell me how to run this code?