Self-Supervised Multimodal Fusion Transformer for Passive Activity Recognition.
Armand K. KoupaiMohammud Junaid BocusRaúl Santos-RodríguezRobert J. PiechockiRyan McConvillePublished in: CoRR (2022)
Keyphrases
- activity recognition
- multimodal fusion
- audio visual
- high robustness
- relevance feedback
- human activity recognition
- multimodal interfaces
- action recognition
- human activities
- daily life
- gait recognition
- accelerometer data
- multi modal
- transfer learning
- data sets
- learning algorithm
- visual information
- eye movements
- object detection
- multimodal interaction
- active learning
- user interface
- spatio temporal
- feature extraction