Awesome

NOTE

the original official VidVRD-II has been deleted, we provide a backup version

refer to vIoU calculation in evaluation/common.py, it has w=xmax-xmin+1 (Line 102), so the coords should not be normalized. I.e., we shall use normalize_coords=False

VidVRD-II

This repository contains source codes for "Video Visual Relation Detection via Iterative Inference" (MM'21) [paper].

Environment

1. Download ImageNet-VidVRD dataset and VidOR dataset. Then, place the data under the same parent folder as this repository (recommended):

├── vidor-dataset
│   ├── annotation
│   │   ├── training
│   │   └── validation
│   └── video
├── imagenet-vidvrd-dataset
│   ├── test
│   ├── train
│   └── videos
├── VidVRD-II
│   ├── ... (this repo)
├── vidor-baseline-output
│   ├── ... (intermediate results)
└── imagenet-vidvrd-baseline-output
    ├── ... (intermediate results)

2. Install dependencies (tested with TITAN Xp GPU)

conda create -n vidvrd-ii -c conda-forge python=3.7 Cython tqdm scipy "h5py>=2.9=mpi*" ffmpeg=3.4 cudatoolkit=10.1 cudnn "pytorch>=1.7.0=cuda101*" "tensorflow>=2.0.0=gpu*"
conda activate vidvrd-ii
python setup.py build_ext --inplace

3. [Optional] Since cv2 is incompatible with the main environment in vidvrd-ii, if you want to use the script visualize.py, you may need to create a separate environment with py-opencv installed.

Quick Start

1. Download the precomputed object tracklets and features for ImageNet-VidVRD (437MB) and VidOR (32GB: part1, part2, part3, part4), and extract them under imagenet-vidvrd-baseline-output and vidor-baseline-output as above, respectively.
2. Run python main.py --cfg config/imagenet_vidvrd_3step_prop_wd0.01.json --id 3step_prop_wd0.01 --train --cuda to train the model for ImageNet-VidVRD. Use --cfg config/vidor_3step_prop_wd1.json for VidOR.
3. Run python main.py --cfg config/imagenet_vidvrd_3step_prop_wd0.01.json --id 3step_prop_wd0.01 --detect --cuda to detect video relations (inference) and the results will be output to ../imagenet-vidvrd-baseline-output/models/3step_prop_wd0.01/video_relations.json.
4. Run python evaluate.py imagenet-vidvrd test relation ../imagenet-vidvrd-baseline-output/models/3step_prop_wd0.01/video_relations.json to evaluate the results.
5. To visualize the results, add the option --visualize to the above command (this will involve visualize.py so please make sure the environment is switched according to the last section). For the better visualization mentioned in the paper, change association_algorithm to graph in the configuration json, and then run Step 3 and 5.
6. To automatically run the whole traininng and test pipepine multiple times, run python main.py --cfg config/imagenet_vidvrd_3step_prop_wd0.01.json --id 3step_prop_wd0.01 --pipeline 5 --cuda --no_cache and then you can obtain a mean/std result.

Object Tracklet Extraction (optional)

1. We extract frame-level object proposals using the off-the-shelf tool. Please first download and install tensorflow model library. Then, run python -m video_object_detection.tfmodel_image_detection [imagenet-vidvrd/vidor] [train/test/training/validation]. You can also download our precomputed results for ImageNet-VidVRD (6GB).
2. To obtain object tracklets based on the frame-level proposals, run python -m video_object_detection.object_tracklet_proposal [imagenet-vidvrd/vidor] [train/test/training/validation].

Acknowledgement

This repository is built based on VidVRD-helper. If this repo is helpful in your research, you can use the following bibtex to cite the paper:

@inproceedings{shang2021video,
  title={Video Visual Relation Detection via Iterative Inference},
  author={Shang, Xindi and Li, Yicong and Xiao, Junbin and Ji, Wei and Chua, Tat-Seng},
  booktitle={Proceedings of the 29th ACM International Conference on Multimedia},
  pages={3654--3663},
  year={2021}
}