Which of the following layer used in PointNet makes it order-invariant to the * 1 point input point cloud data? Convolutional layers Global max pooling layer Fully connected layers None of them. It's something else PointNet is not order-invariant Which one of the following options can be used to represent 3D input data * 1 point in deep learning? Point cloud Mesh Voxel All of them Which ones of the following statements regarding segmentation are * 1 point correct? (A) Semantic segmentation is a pixel-wise classification problem (B) Instance segmentation can be solved by semantic segmentation networks (C) Panoptic segmentation is semantic segmentation plus instance segmentation (D) Mask-RCNN has a similar network architecture as Faster RCNN (A), (B) (D) (A), (C), (D) (C), (D) (B), (C)
