ULSAM: Ultra-Lightweight Subspace Attention Module for Compact Convolutional Neural Networks.
Rajat SainiNandan Kumar JhaBedanta DasSparsh MittalC. Krishna MohanPublished in: WACV (2020)
Keyphrases
- lightweight
- convolutional neural networks
- convolutional network
- visual attention
- dimensionality reduction
- high speed
- low dimensional
- principal component analysis
- subspace clustering
- feature space
- wireless sensor networks
- communication infrastructure
- feature extraction
- data driven
- development environments
- handheld devices
- rfid tags
- high dimensional data
- dynamic programming
- high dimensional
- multiscale