Skip to content

imnotk/action_tf

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Action-tf

This is a implementation of CCF-Net in TensorFlow.

Requirement

  • tensorflow 1.4
  • dm-sonnet
  • Opencv-python

Dataset preparation

For ucf101 and hmdb51, we recommend download from origin webset

For Optical flow, we recommend download from twostreamfusion, or you can generate your own optical flow dataset by following the TSN setting

Training and Testing

To train a new model, all my training setting is under the root of "script".

bash ts_rgb_resnet50.sh
bash ts_flow_resnet50.sh
bas_resnet50_non_local_fusion.sh

Eval Score

Use the following command to test its performance of ucf101:

python eval_score_resnet50.py -modality fusion --dataset UCF101

Minor difference

原始TSN使用pytorch和caffer实现,由于用tensorflow实现可能会存在一定的性能差异,以下时split1上的性能

Modality ResNet50 ResNet101
RGB 84.6%~84.8% 86.3%~87.2%
Flow 87.2%~87.4% 88.2%~88.4%
CCF-Net 93.6%~93.8% 94.4%~94.6%

confusion matrix

每次使用train.py和test.py后会在相应的文件夹中生成准确率等on-the-fly文件如:

logdir/UCF101/0/TS_resnet50/rgb/video_predict_multi.pickle
logdir/UCF101/0/TS_resnet50/rgb/train_log.txt
logdir/UCF101/0/TS_resnet50/rgb/val_log.txt
logdir/UCF101/0/TS_resnet50/rgb/test_result_multi.txt
logdir/UCF101/0/TS_resnet50/rgb/test_result_center.txt

随后生成混淆矩阵

python ./script/confusion_matrix_ts_resnet50.py

重要通知

由于tensorflow1.x没有自带的ImageNet预训练权重,我们需要从Tensorflow Models里面下载slim的模型,再转换成dm-sonnet的模型,需要你先从slim中下载ResNet50等模型权重(不要下载成TF2的),然后使用utils目录下的rebuild_ckpt.py文件转换模型。 注意:我没有提供argparser等命令行指令,需要你手动修改你的模型权重路径。

python rebuild_ckpt.py

对于数据集同理,由于没有提供整体的config文件,你需要在Dataset里面调整各个文件的路径,这只是个Draft文件。

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published