[PRCV-2023] Learning Bottleneck Transformer for Event Image-Voxel Feature Fusion based Classification, Chengguo Yuan, Yu Jin, Zongzhen Wu, Fanting Wei, Yangzirui Wang, Lan Chen, and Xiao Wang. [arXiv] [Paper_with_Code] [Poster]
- Check the source code of EFV++ (Journal extension, will release the code soon!!!)
Recognizing target objects using an event-based camera draws more and more attention in recent years. Existing works usually represent the event streams into point-cloud, voxel, image, etc, and learn the feature representations using various deep neural networks. Their final results may be limited by the following factors: monotonous modal expressions and the design of the network structure. To address the aforementioned challenges, this paper proposes a novel dual-stream framework for event representation, extraction, and fusion. This framework simultaneously models two common representations: event images and event voxels. By utilizing Transformer and Structured Graph Neural Network (GNN) architectures, spatial information and three-dimensional stereo information can be learned separately. Additionally, a bottleneck Transformer is introduced to facilitate the fusion of the dual-stream information. Extensive experiments demonstrate that our proposed framework achieves state-of-the-art performance on two widely used event-based classification datasets.
Python 3.8
Pytorch
numpy
scipy
Pytorch Geometric
torch-cluster 1.5.9
torch-geometric 1.7.0
torch-scatter 2.0.6
torch-sparse 0.6.9
torch-spline-conv 1.2.1
spconv
pandas
pillow
Matlab
ASL-DVS:https: //www.dropbox.com/sh/ibq0jsicatn7l6r/AACNrNELV56rs1YInMWUs9CAa?dl=0
N-MNIST: https://www.garrickorchard.com/datasets/n-mnist
python downsample/to_voxel.py
python generate_graph/voxel2graph.py
train
CUDA_VISIBLE_DEVICES=0 python -m torch.distributed.launch --master_port 1568 --nproc_per_node=1 train.py --epoch 150 --batch_size 8
test
CUDA_VISIBLE_DEVICES=0 python -m torch.distributed.launch --nproc_per_node=1 test.py --model_name xxx.pkl --batch_size 8
If you think this work contributes to your research, please give us a star 🌟 and cite this paper via:
@misc{yuan2023learning,
title={Learning Bottleneck Transformer for Event Image-Voxel Feature Fusion based Classification},
author={Chengguo Yuan and Yu Jin and Zongzhen Wu and Fanting Wei and Yangzirui Wang and Lan Chen and Xiao Wang},
year={2023},
eprint={2308.11937},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
If you have any questions about this work, please leave an issue.