An introduction to factor graphs

By Hans-Andrea Loeliger. IEEE Signal Processing Magazine, 2004

988 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Networks and Communications, Electrical and Electronic Engineering

View Paper: https://www.semanticscholar.org/paper/f054792ab3ec8d827d84d007dc2bd01e69c4cea7

How researchers see this paper

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

In Artificial Intelligence

In Computer Networks and Communications

In Electrical and Electronic Engineering

Citations in the corpus

Collaborative Multiagent Reinforcement Learning by Payoff Propagation (2006, background): “Furthermore, it is possible to generalize the proposed techniques to payoff functions with more than two agents because any arbitrary graph can be converted to a graph with only pairwise inter-agent dependencies (Yedidia et al., 2003; Loeliger, 2004).”
Graphical Models, Exponential Families, and Variational Inference (2008, background): “The formalism of factor graphs provides an alternative graphical representation, one which emphasizes the factorization of the distribution [140, 154].”
D3VO: Deep Depth, Deep Pose and Deep Uncertainty for Monocular Visual Odometry (2020, methodology): “Unlike traditional direct VO approaches [19,23] which initialize the front-end tracking for each new frame with a constant velocity motion model, we leverage the predicted poses between consecutive frames to build a non-linear factor graph [41, 47].”
Collective multi-label classification (2005, methodology): “Inference in factor graphs is done in a way similar to inference in graphical models [10].”
Message Passing Algorithms for Scalable Multitarget Tracking (2018, positive): “The factorization (4) can be represented by a factor graph [25], [28].”
Channel Estimation for OFDM (2014, methodology): “Although EXIT chart or density evolution can be used to explain the performance of the receiver in [103], [105], [165] and [166], there is still no analytical explanation on the behavior of the factor graph receiver when there are loops in the graph [115], [126].”
A survey of uncertainty in deep neural networks (2021, methodology): “As a result, many probabilistic methods such as factor graphs [308], [309] are now the work-horse of advanced consumer products such”
Model-Based Deep Learning (2020, background): “When the block size t is large, the messages may tend to zero, and are thus commonly scaled [84], e.”
Simulation-Based Computation of Information Rates for Channels With Memory (2006, background): “(This graph is a Forney-style factor graph, see [16], [28]; add a circle on each branch to obtain a factor graph as in [26].”
Power systems data fusion based on belief propagation (2017, background): “A brief introduction of factor graphs is provided in section III-A (see [15]–[17] for more details).”
Massive MIMO Detection Techniques: A Survey (2019, methodology): “This property can be exploited to model the MIMO system by a factor graph (Tanner graph) [117] as presented in Fig.”
Exploiting Agent and Type Independence in Collaborative Graphical Bayesian Games (2011, background): “Fortunately, an interaction hyper-graph can be translated into afactor graph(Kschischang et al., 2001; Loeliger, 2004) to which MAX -”
Cooperative Localization in Wireless Networks (2009, methodology): “The localization process typically consists of two phases [ 1 ].”
Distributed learning in wireless sensor networks (2005, background): “This surge in popularity is inspired in part by the powerful graphical model framework that has enabled many exciting applications and inspired new theoretical tools [1], [40], [46], [53], [70], [73], [74].”
The Factor Graph Approach to Model-Based Signal Processing (2007, methodology): “(Formally, a B-valued known variable is an element of B while a B-valued unknown variable is a function from the configuration space into B [ 3 ].) For the factor graph of fðxj y1; y2Þ ,w e would thus modify Fig. 2b y replacingY1 by y2 and Y2 by y2 if these variables are known (observed).”