SuPEr - SAM: Using the Supervision Signal from a Pose Estimator to Train a Spatial Attention Module for Personal Protective Equipment Recognition.
Adrian SandruGeorgian-Emilian DutaMariana-Iuliana GeorgescuRadu Tudor IonescuPublished in: WACV (2021)
Keyphrases
- visual focus of attention
- focus of attention
- robust recognition
- pose invariant face recognition
- object recognition
- recognition rate
- improve recognition accuracy
- spatial filtering
- spatio temporal
- face recognition under varying
- pattern recognition
- spatial and temporal
- least squares
- pose estimation
- illumination invariance
- recognition accuracy
- object representation
- power line
- non stationary
- partial occlusion
- spatial data
- recognize faces
- recognition algorithm
- signal processing
- personal information
- maximum likelihood
- visual attention
- frequency domain
- feature extraction
- noisy environments
- character recognition
- human activities
- activity recognition
- high frequency
- action recognition
- head pose estimation
- image matching
- wavelet transform
- high resolution
- human body