ST-DETR: Spatio-Temporal Object Traces Attention Detection Transformer
Eslam Mohamed BAKR Ahmad El-Sallab
Abstract
We propose ST-DETR, a Spatio-Temporal Transformer-based architecture for object detection from a sequence of temporal frames. We treat the temporal frames as sequences in both space and time and employ the full attention mechanisms to take advantage of the features correlations over both dimensions. This treatment enables us to deal with frames sequence as temporal object features traces over every location in the space. We explore two possible approaches; the early spatial features aggregation over the temporal dimension, and the late temporal aggregation of object query spatial features. Moreover, we propose a novel Temporal Positional Embedding technique to encode the time sequence information. To evaluate our approach, we choose the Moving Object Detection (MOD) task, since it is a perfect candidate to showcase the importance of the temporal dimension. Results show a significant 5% mAP improvement on the KITTI MOD dataset over the 1-step spatial baseline
Poster
Citation
@misc{mohamed2021stdetr,
title={ST-DETR: Spatio-Temporal Object Traces Attention Detection Transformer},
author={Eslam Mohamed and Ahmad El-Sallab},
year={2021},
eprint={2107.05887},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
Authors
Eslam Mohamed BAKR
Senior Machine Learning Engineer at Valeo
Ahmad El-Sallab
AI Senior Expert at Valeo
Questions?
eslam.mohamed-abdelrahman@valeo.com