All Fingers Are Not the Same: Handling Variable-Length Sequences in a Discriminative Setting Using Conformal Multi-Instance Kernels.
Sarvesh NikumbhPeter EbertNico PfeiferPublished in: WABI (2017)
Keyphrases
- variable length
- multi instance
- fixed length
- real valued
- multi instance learning
- multi label
- semi supervised learning
- n gram
- semi supervised
- bitstream
- kernel methods
- feature space
- object segmentation
- support vector
- kernel function
- unsupervised learning
- object detection
- training data
- metric learning
- feature extraction
- text categorization
- machine learning
- hidden markov models
- high dimensional
- human motion
- object recognition
- face recognition
- image processing
- feature selection