SPIN: A High Speed, High Resolution Vision Dataset for Tracking and Action Recognition in Ping Pong.
Steven SchwarczPeng XuDavid D'AmbrosioJuhana KangaspuntaAnelia AngelovaHuong PhanNavdeep JaitlyPublished in: CoRR (2019)
Keyphrases
- action recognition
- human actions
- high speed
- video dataset
- ping pong
- high resolution
- human detection
- computer vision
- recognition of human actions
- real time
- depth sensors
- body parts
- human pose
- activity recognition
- robust tracking
- low resolution
- ucf sports
- bag of words
- image processing
- action classification
- object tracking
- particle filter
- action recognition in videos
- space time interest points
- recognizing actions
- super resolution
- recognizing human actions
- bag of features
- vision system
- static images
- human activities
- visual tracking
- space time
- action detection
- human motion
- view invariant
- depth cameras
- partial occlusion
- kalman filter
- three dimensional
- action primitives
- appearance model
- visual features
- face images
- high quality
- machine learning