About evaluation indicators

ttengwang / PDVC

End-to-End Dense Video Captioning with Parallel Decoding (ICCV 2021)

MIT License

200 stars 23 forks source link

About evaluation indicators #34

Closed llljjj88 closed 2 years ago

llljjj88 commented 2 years ago

I would like to ask if the evaluation index of the reproduction model of the paper is only described by video paragraphs. If I want to get the evaluation index of dense video caption

ttengwang commented 2 years ago

Sorry, I can't understand the question. Could you explain it a little bit more?

llljjj88 commented 2 years ago

您好，那个关于评价指标的问题就是只有para_bleu的指标但是没有关于dense video caption的评价指标bleu的，那么我将如何才能得到这些指标呢，另外我复现的模型训练速度太久了，我可以修改哪些参数使得训练速度可以提升呢？感谢

ttengwang commented 2 years ago

模型可以预测METEOR，Recall，Precision，soda_c等指标, 这些都是dense video captioning的指标. 如果想得到BLEU和Cider, 将这一行修改为args.verbose = True, 修改后运行速度会慢一些. https://github.com/ttengwang/PDVC/blob/7d4f3774516f17ab5a676dec0234c15780e8eaa5/densevid_eval3/eval_dvc.py#L9

ttengwang commented 2 years ago

V100上一个epoch时间在半个小时内, epoch设置为25左右, 一次实验是半天, 应该不会太久

llljjj88 commented 2 years ago

谢谢您的回复解决了我的困惑