Improved adaptive Gaussian mixture model for background subtraction

By Zoran Zivkovic. Proceedings of the 17th International Conference on Pattern Recognition, 2004. ICPR 2004., 2004

2,590 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/d44d0acccb25fb0a3cb86bdba066039145df1a36

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

Automatic HVAC Control with Real-time Occupancy Recognition and Simulation-guided Model Predictive Control in Low-cost Embedded System (2017, methodology): “To overcome these challenges, we employ the Gaussian mixture-based background/foreground segmentation algorithm proposed in [39, 40], and implemented on OpenCV.”
ToxTrac: A fast and robust software for tracking organisms (2017, background): “8: Threshold, this parameter selects the minimum intensity level (1-255) for the detected objects.”
Comparative study of background subtraction algorithms (2010, methodology): “For example, in [21] and [35] An unstructured approach can also be used to model a multimodal PDF.”
Counting People With Low-Level Features and Bayesian Regression (2012, methodology): “As an alternative to motion segmentation, two background subtraction methods, i.e., a temporal median filter and an adaptive GMM [45], were used to obtain the “scene” segment, which was then used for count regression.”
A texture-based method for modeling the background and detecting moving objects (2006, background): “This is maybe due to the high-computational complexity and limited performance of texture methods.”
Recent Advanced Statistical Background Modeling for Foreground Detection - A Systematic Survey (2011, background): “Background Maintenance Variable K Zivkovic [49], Cheng et al.”
Background segmentation with feedback: The Pixel-Based Adaptive Segmenter (2012, neutral): “GMM— Zivkovic [10] 0.”
Introduction to the special topic section on mining Web resources for enhancing information retrieval (2007, methodology): “Huo, Chan, and Lee (1995) proposed a Bayesian adaptive algorithm for learning the parameters in the hidden Markov model for speech recognition Zivkovic (2004) developed an improved method to reestimate recursively the parameters in Gaussian mixture models for background subtraction, which is a com”
Rain Removal in Traffic Surveillance: Does it Matter? (2018, methodology): “Furthermore, we include the Mixture of Gaussians (MoG) method as modified by Zivkovic [77] as this is a classic segmentation method that is well understood and often used as a baseline for comparisons.”
Human activities recognition using depth images (2013, background): “There is an emerging consensus that depth information is a more reliable recognition cue for classifying human activities.”
Background Modeling using Mixture of Gaussians for Foreground Detection - A Survey (2008, methodology): “[94] Estimation at each pixel - Dirichlet prior yes IIR Filter”
SuBSENSE: A Universal Change Detection Method With Local Adaptive Sensitivity (2015, background): “New adaptive and more flexible variations of GMM were also proposed over the years [10]–[13] to allow dynamic numbers of components for modeling as well as to improve their convergence rate.”
Camera-trap images segmentation using multi-layer robust principal component analysis (2017, methodology): “tial coherence-based approach (Fuzzy-Adaptive-SOM) [21], Gaussian Mixture Model (GMM) [34], Adaptive Gaussian Mixture Model (AGMM) [43], Texture-based method (DPTexture) [13], and multi-layer background subtraction (multi-layer) [39].”
Focus: Querying Large Video Datasets with Low Latency and Low Cost (2018, methodology): “chines, with each machine running one worker process for each video stream’s ingestion. The ingest worker receives the live video stream, and extracts the moving objects (using background subtraction [81]); it is extensible to plug in any other object detector. The detected objects are sent to the i”
Variational Deep Embedding: An Unsupervised and Generative Approach to Clustering (2016, methodology): “ optimized by the Expectation Maximization (EM) algorithm. One advantage of GMM over K-means is that a GMM can generate samples by estimation of data density. Although K-means, GMM and their variants [33,27,25,38,15,8] have been extensively used, learning good representations most suitable for clust”
A hand pose tracking benchmark from stereo matching (2017, methodology): “B5 in Fig.4 (b) shows that the hand probability from the proposed method is relatively high in the background since its background also has fast motion objects and adaptive GMM treats them as foreground.”