Subgraph decomposition for multi-target tracking

By Siyu Tang, Bjoern Andres, Mykhaylo Andriluka, Bernt Schiele. 2015

188 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

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

Citations in the corpus

Learning Non-maximum Suppression (2017, methodology): “Principled clustering formulations with globally optimal solutions have been proposed in [27, 23], although they have yet to surpass the performance of GreedyNMS.”
Joint Graph Decomposition & Node Labeling: Problem, Algorithms, Applications (2016, background): “[14] introduce a binary linear program w.”
Features for Multi-target Multi-camera Tracking and Re-identification (2018, background): “Methods that use all pairwise terms, not only timeconsecutive ones, are significantly more accurate but NPhard [18, 25, 27, 28, 41, 58, 61, 65, 66, 67].”
Joint Graph Decomposition and Node Labeling by Local Search (2016, background): “Conceptually, the problem we propose generalizes the unconstrained integer quadratic program and the minimum cost lifted multicut problem, both of which areNP-hard.”
Online Multi-Object Tracking with Dual Matching Attention Networks (2018, methodology): “Typically, these methods model the MOT task as a global optimization problem in various forms such as network ﬂow [58,51,14], and multi-cut [47,48,46].”
MOTChallenge: A Benchmark for Single-Camera Multiple Target Tracking (2020, methodology): “targets and data association simultaneously, reaching 30.72% MOTA on MOT15 (GMPHD OGM), 41% and 40.42% on MOT16 (PHD GSDL and GMPHD ReId, respectively) and 49.94% (GMPHD OGM) on MOT17. Newer methods [Tang et al., 2015] bypassed the need to pre-process object detections with NMS. They proposed a mult”
UA-DETRAC: A new benchmark and protocol for multi-object detection and tracking (2015, background): “The sequences of three levels of di ﬃ culties, i.e. , easy , medium , and hard are denoted in purple, blue, and orange, respectively.”
A Convnet for Non-maximum Suppression (2015, methodology): “In our base Tnet the first stage applies 512 11×11 filters over each input layer, and 512 1 × 1 filters are applied on layers 2, 3, and 4.”
Deep Affinity Network for Multiple Object Tracking (2018, unknown): “non-maximum suppression for tracking [48], [49].”
Multiple object tracking: A literature review (2014, methodology): “boolean programming [129], K-shortest paths [18], [42], set cover [130] and subgraph multicut [131], [132] are adopted to solve the association problem among detections or tracklets.”
Spatial-Temporal Relation Networks for Multi-Object Tracking (2019, background): “According to what kind of information is used to establishing the association between objects in different frames, the existing methods can be categorized into online methods [2, 12, 19, 27, 33, 35, 38, 47, 50, 58, 59, 9], and offline methods [7, 14, 36, 39, 40, 44, 51, 52, 53, 62].”
Online Multi-object Tracking Using CNN-Based Single Object Tracker with Spatial-Temporal Attention Mechanism (2017, methodology): “These off-line methods consider MOT as a global optimization problem and focus on designing various optimization algorithm such as network flow [39, 58], continuous energy minimization [33], max weight independent set [6], k-partite graph [56, 10], subgraph multi-cut [43, 44] and so on.”
ArtTrack: Articulated Multi-Person Tracking in the Wild (2016, methodology): “We use the solver from [20] for consistency, but a simpler KL-based solver as in [25, 19] could be used as well since the TD/BU model effectively ignores the unary variables xd.”
Learning a Neural Solver for Multiple Object Tracking (2019, background): “non-maxima suppressed) detections to approach this challenge [9, 25, 24, 26, 12].”
Multiple People Tracking by Lifted Multicut and Person Re-identification (2017, background): “[27, 28] propose to jointly cluster detections over space and time.”