DeepPBM: Deep Probabilistic Background Model Estimation from Video Sequences.
Behnaz RezaeiAmirreza FarnooshSarah OstadabbasPublished in: ICPR Workshops (2) (2020)
Keyphrases
- background model
- video sequences
- foreground objects
- background subtraction
- moving objects
- dynamic background
- background modeling
- moving object detection
- moving camera
- motion detection
- pixel level
- image sequences
- object tracking
- probabilistic model
- video surveillance
- gaussian mixture model
- dynamic scenes
- video frames
- template matching
- frame rate
- segmentation algorithm
- moving target
- object segmentation
- pixel wise
- motion vectors
- space time
- video camera
- video data
- bayesian networks
- motion estimation
- camera motion
- object detection
- appearance model
- pixel values
- human actions
- real time
- post processing
- mixture model
- model selection
- computer vision