A Vision-based Human Action Recognition System for Moving Cameras Through Deep Learning.
Ming-Jen ChangJih-Tang HsiehChiung-Yao FangSei-Wang ChenPublished in: SPML (2019)
Keyphrases
- deep learning
- human actions
- moving camera
- video sequences
- action recognition
- space time
- human motion
- dynamic scenes
- spatio temporal
- image sequences
- camera motion
- moving objects
- d scene
- unsupervised learning
- machine learning
- video camera
- field of view
- human computer interaction
- background subtraction
- computer vision
- activity recognition
- optical flow
- object tracking
- human activities
- feature extraction
- video data
- motion estimation
- mental models
- image processing
- image classification
- active learning
- high resolution