search

MarcusNerva / HMN

[CVPR2022] Official code for Hierarchical Modular Network for Video Captioning. Our proposed HMN is implemented with PyTorch.
MIT License
52 stars 9 forks source link

Dataset processing #32

Open fsh2017 opened 9 months ago

fsh2017 commented 9 months ago

I want to train my own dataset, and I'm unsure about how many of the text features are generated. Could you provide some specific guidance on this? Thank you very much!