Deep Multimodal Feature Encoding for Video Ordering

Paper

Vivek Sharma, Makarand Tapaswi, and Rainer Stiefelhagen

In IEEE International Conference on Computer Vision (ICCV) workshop on Large Scale Holistic Video Understanding, 2019

Temporal Compact Bilinear Pooling (TCBP) Layer

Demo:

$ from TCBP import TCBP
$ import torch
$ data = torch.rand([10,8192,4,1,1]) 
$ tcbp = TCBP(input_dim1=8192, input_dim2=8192,output_dim=512, temporal_window=4, spat_x=1, spat_y=1)
$ tcbp_representation = tcbp(data,data)
$ tcbp_representation.shape  
$ ---> torch.Size([10, 512])

Citation

If you find the code and datasets useful in your research, please cite:

@inproceedings{tcbp,
    author    = {Sharma, Vivek and Tapaswi, Makarand and Stiefelhagen, Rainer}, 
    title     = {Deep Multimodal Feature Encoding for Video Ordering}, 
    booktitle = {IEEE ICCV Workshop on Large Scale Holistic Video Understanding},
    year      = {2019}
}