@article{Weber2021,
   abstract = {For reliable environment perception, the use of temporal information is
essential in some situations. Especially for object detection, sometimes a
situation can only be understood in the right perspective through temporal
information. Since image-based object detectors are currently based almost
exclusively on CNN architectures, an extension of their feature extraction with
temporal features seems promising. Within this work we investigate different architectural components for a
CNN-based temporal information extraction. We present a Temporal Feature
Network which is based on the insights gained from our architectural
investigations. This network is trained from scratch without any ImageNet
information based pre-training as these images are not available with temporal
information. The object detector based on this network is evaluated against the
non-temporal counterpart as baseline and achieves competitive results in an
evaluation on the KITTI object detection dataset.},
   author = {Michael Weber and Tassilo Wald and J. Marius Zollner},
   doi = {10.1109/IV48863.2021.9575941},
   isbn = {9781728153940},
   journal = {IEEE Intelligent Vehicles Symposium, Proceedings},
   month = {3},
   pages = {1478-1484},
   publisher = {Institute of Electrical and Electronics Engineers Inc.},
   title = {Temporal Feature Networks for CNN based Object Detection},
   volume = {2021-July},
   url = {https://arxiv.org/abs/2103.12213v1},
   year = {2021},
}