Awesome

NVI-DEHR

Code for our ECCV 2024 paper "Nonverbal interaction detection".

Contributed by Jianan Wei*, Tianfei Zhou*, Yi Yang, and Wenguan Wang.

Installation

Install the dependencies.

conda create -n NVI python=3.8
python -m pip install --upgrade pip==23.3.1
pip install -r requirements_lavis.txt
git clone https://github.com/salesforce/LAVIS.git
cd LAVIS
pip install -e .
pip install -r requirements.txt
pip install torch==1.12.1+cu113 torchvision==0.13.1+cu113 torchaudio==0.12.1 --extra-index-url https://download.pytorch.org/whl/cu113

Some files in the LAVIS library have been modified to extract complete visual features from the CLIP model.

Data preparation

NVI

NVI dataset can be downloaded here. The downloaded annotation files have to be placed as follows.

data
 └─ NVI
     |─ annotations
     |   |─ train.json
     |   |─ test.json
     |   └─ val.json
     |─ images
     :

HOI-DET

HICO-DET dataset can be downloaded here and the annotation files can be downloaded from here.

V-COCO dataset can be downloaded here and the prior file prior.pickle from here.

For more details, check out the GEN-VLKT.

Pre-trained model

Download the pre-trained model of DETR detector for ResNet50, and put it in the params directory.

python ./tools/convert_parameters.py \
        --load_path params/detr-r50-e632da11.pth \
        --save_path params/detr-r50-pre-2branch-nvi.pth \
        --num_queries 64

python ./tools/convert_parameters.py \
        --load_path params/detr-r50-e632da11.pth \
        --save_path params/detr-r50-pre-2branch-hico.pth \
        --num_queries 64

python ./tools/convert_parameters.py \
        --load_path params/detr-r50-e632da11.pth \
        --save_path params/detr-r50-pre-2branch-vcoco.pth \
        --dataset vcoco \
        --num_queries 64

Training

After the preparation, you can start training with the following commands.

NVI

sh ./config/nvi.sh

HICO-DET

sh ./config/hico.sh

V-COCO

sh ./configs/vcoco.sh

Evaluation

NVI

You can evaluate with trained parameters for NVI as follows.

python -m torch.distributed.launch \
        --nproc_per_node=4 \
        --use_env \
        main.py \
        --pretrained pretrained/detr-r50-pre-2branch-nvi.pth \
        --dataset_file vpic \
        --backbone resnet50 \
        --num_queries 64 \
        --dec_layers 3 \
        --num_workers 2 \
        --epochs 90 \
        --lr_drop 60 \
        --batch_size 4 \
        --use_nms_filter \
        --set_cost_bbox 2.5 \
        --set_cost_giou 1 \
        --bbox_loss_coef 2.5 \
        --giou_loss_coef 1 \
        --mimic_loss_coef 2 \
        --n_layer 5 \

HICO-DET

You can conduct the evaluation with trained parameters for HICO-DET as follows.

python -m torch.distributed.launch \
        --nproc_per_node=8 \
        --use_env \
        main.py \
        --pretrained pretrained/hico_gen_vlkt_s.pth \
        --dataset_file hico \
        --hoi_path data/hico_20160224_det \
        --num_obj_classes 80 \
        --num_verb_classes 117 \
        --backbone resnet50 \
        --num_queries 64 \
        --dec_layers 3 \
        --eval \
        --with_clip_label \
        --with_obj_clip_label \
        --use_nms_filter

For the official evaluation (reported in paper), you need to covert the prediction file to a official prediction format following this file, and then follow PPDM evaluation steps.

V-COCO

Firstly, you need the add the following main function to the vsrl_eval.py in data/v-coco.

if __name__ == '__main__':
  import sys

  vsrl_annot_file = 'data/vcoco/vcoco_test.json'
  coco_file = 'data/instances_vcoco_all_2014.json'
  split_file = 'data/splits/vcoco_test.ids'

  vcocoeval = VCOCOeval(vsrl_annot_file, coco_file, split_file)

  det_file = sys.argv[1]
  vcocoeval._do_eval(det_file, ovr_thresh=0.5)

Next, for the official evaluation of V-COCO, a pickle file of detection results has to be generated. You can generate the file with the following command. and then evaluate it as follows.

python generate_vcoco_official.py \
        --param_path pretrained/VCOCO_GEN_VLKT_S.pth \
        --save_path vcoco.pickle \
        --hoi_path data/v-coco \
        --num_queries 64 \
        --dec_layers 3 \
        --use_nms_filter \
        --with_clip_label \
        --with_obj_clip_label

cd data/v-coco
python vsrl_eval.py vcoco.pickle

Citation

Please consider citing our paper if it helps your research.

@article{wei2024nonverbal,
  title={Nonverbal Interaction Detection},
  author={Wei, Jianan and Zhou, Tianfei and Yang, Yi and Wang, Wenguan},
  journal={arXiv preprint arXiv:2407.08133},
  year={2024}
}

License

NVI-DEHR is released under the MIT license. See LICENSE for additional details.

Acknowledge

Some of the codes are built upon PPDM, DETR, QPIC and CDN, GEN-VLKT. Thanks for their great work!