G2L: A High-Dimensional Geometric Approach for Automatic Generation of Highly Accurate Pseudo-labels.
John R. KenderParijat DubeZhengyang HanBishwaranjan BhattacharjeePublished in: ICCV (Workshops) (2023)
Keyphrases
- highly accurate
- high dimensional
- capable of producing
- high quality
- similarity search
- high dimensionality
- high dimensional data
- automatically generate
- low dimensional
- high accuracy
- dimensionality reduction
- feature space
- multi label
- pairwise
- sparse data
- data points
- accurate models
- multi modal
- geometric information
- geometric structure
- variable selection
- nearest neighbor
- parameter space
- multiscale
- neural network
- multi dimensional
- sparse coding
- metric space
- manifold learning
- training data
- class labels
- decision trees
- kernel function
- automatically generating
- machine learning