Deep learning-based object detection methods have achieved great performance improvement. However, since small kernel convolution has been widely used, the semantic feature is difficult to obtain due ...
To avoid the problem of the traditional methods consuming large computational resources to calculate the kernel matrix and 2D discrete convolution, we present a novel approach for 3D gravity and ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results