video-style-transfer's Introduction

video-style-transfer

This is video style transfer PyTorch implementation based on "Arbitrary Style Transfer with Style-Attentional Networks".

Official paper: https://arxiv.org/abs/1812.02342v5.

Source code: https://github.com/GlebBrykin/SANET

Dataset:

COCO
WikiArt
Video sequence（60 videos， from https://www.videvo.net/）

Modify:

Add temporal loss and Spatial smoothing loss to fine-tune.
Use image pair from video to fine-tune.

Train:

Image_train: COCO+WikiArt
video_train: Video sequence+WikiArt

Test:

Image_transfer: single image transfer
video_transfer: video transfer

Result:

demo1:https://pan.baidu.com/s/1o40EPY7_6FnMKsnjGaD24Q
demo2:https://pan.baidu.com/s/1ZMPegXQCBB35NimzmEg2fQ

video-style-transfer's People

Contributors

Stargazers

Watchers

video-style-transfer's Issues

Error when load weight

Can you help me this problem?

Traceback (most recent call last):
  File "image_transfer.py", line 55, in <module>
    decoder.load_state_dict(torch.load(args.decoder))
  File "/home/server2/son_env/lib/python3.6/site-packages/torch/nn/modules/module.py", line 1224, in load_state_dict
    self.__class__.__name__, "\n\t".join(error_msgs)))
RuntimeError: Error(s) in loading state_dict for Decoder:
	Missing key(s) in state_dict: "features.1.weight", "features.1.bias", "features.5.weight", "features.5.bias", "features.8.weight", "features.8.bias", "features.11.weight", "features.11.bias", "features.14.weight", "features.14.bias", "features.18.weight", "features.18.bias", "features.21.weight", "features.21.bias", "features.25.weight", "features.25.bias", "features.28.weight", "features.28.bias". 
	Unexpected key(s) in state_dict: "1.weight", "1.bias", "5.weight", "5.bias", "8.weight", "8.bias", "11.weight", "11.bias", "14.weight", "14.bias", "18.weight", "18.bias", "21.weight", "21.bias", "25.weight", "25.bias", "28.weight", "28.bias".