Spatiality-guided Transformer for 3D Dense Captioning on Point Clouds.
Heng WangChaoyi ZhangJianhui YuTom Weidong CaiPublished in: CoRR (2022)
Keyphrases
- point cloud
- multi view reconstruction
- architectural scenes
- surface reconstruction
- structure from motion
- point sets
- laser scanner
- point cloud data
- scene flow
- stereo correspondence
- computer vision
- photorealistic
- lidar data
- stereo camera
- normal vectors
- triangular mesh
- neural network
- camera motion
- object recognition
- image sequences
- machine learning