-
We provide the latest version of ECO-pytorch and pretrained models here.
-
Many thanks to the author @mzolfaghari.
-
If you have any questions, feel free to open a new issue in this repo.
-
Pre-trained model for 2D-Net is provided by tsn-pytorch.
-
Codes modified from tsn-pytorch.
"ECO: Efficient Convolutional Network for Online Video Understanding"
By Mohammadreza Zolfaghari, Kamaljeet Singh, Thomas Brox
paper link
- Python 3.6.4
- PyTorch 0.3.1
git clone https://github.com/zhang-can/ECO-pytorch
python gen_dataset_lists.py <ucf101/something> <dataset_frames_root_path>
e.g. python gen_dataset_lists.py something ~/dataset/20bn-something-something-v1/
The dataset should be organized as:
<dataset_frames_root_path>/<video_name>/<frame_images>
[UCF101 - ECO - RGB] command:
python main.py ucf101 RGB <ucf101_rgb_train_list> <ucf101_rgb_val_list> \
--arch ECO --num_segments 4 --gd 5 --lr 0.001 --lr_steps 30 60 --epochs 80 \
-b 32 -i 1 -j 1 --dropout 0.8 --snapshot_pref ucf101_ECO --rgb_prefix img_ \
--consensus_type identity --eval-freq 1