Video recognition is an important problem with many potential applications.

To better understand this trade-off, we observe that sparse and low-rank approximations excel in different regimes, determined by the softmax temperature in attention, and sparse + low-rank can outperform each individually.

Detecting 3D objects with multi-modality sensors ( i.e., LiDAR and camera) is regarded as a fundamental task in real-world scenes. For accurate object detection, data from different modalities are utilized to provide complementary knowledge, like accurate positions from point clouds and rich context from images.