Impact of Signal-to-Noise Ratio and Bandwidth on Graph Laplacian Spectrum From High-Dimensional Noisy Point Cloud.
Xiucai DingHau-Tieng WuPublished in: IEEE Trans. Inf. Theory (2023)
Keyphrases
- signal to noise ratio
- point cloud
- graph laplacian
- laplace beltrami
- high dimensional
- noise reduction
- wiener filter
- structure from motion
- random walk
- spectral clustering
- low dimensional
- euclidean space
- point sets
- spectral analysis
- basis functions
- similarity search
- kernel machines
- edge detection
- metric space
- pointwise
- manifold structure
- high dimensional data
- dimensionality reduction
- weighted graph
- sparse coding
- image processing
- missing data
- nearest neighbor
- feature space
- kernel function
- labeled and unlabeled data
- neighborhood graph
- embedding space
- training data