Multiple Hypothesis Tracking Revisited

Kim, Chanho; Li, Fuxin; Ciptadi, Arridhana; Rehg, James M.

doi:10.1109/iccv.2015.533

Cited by 603 publications

(444 citation statements)

References 33 publications

Supporting

Mentioning

443

Contrasting

Unclassified

Order By: Relevance

“…Algorithms representing the former category, such as Global Nearest Neighbor (GNN) and Joint Probabilistic Data Association (JPDA) filtering, integer programming or Multi Hypothesis Tracking (MHT) are well-described in Bar-Shalom, Daum, and Huang (2009), Kim, Li, Ciptadi, and Rehg (2015), Pulford (2005) and Khaleghi, Khamis, Karray, and Razavi (2013). More recent work has also applied Random Finite Set (RFS) theory to MTT yielding the Probability Hypothesis Density (PHD) or Cardinalized PHD (CPHD) filters Mahler (2015).…”

Section: Multi-target Trackersmentioning

confidence: 99%

Validation of Radar Image Tracking Algorithms with Simulated Data

Heymann¹,

Hoth²,

Banyś³

et al. 2017

TransNav

View full text Add to dashboard Cite

Collision avoidance is one of the high-level safety objectives and requires a complete and reliable description of the maritime traffic situation. The radar is specified by the IMO as the primary sensor for collision avoidance. In this paper we study the performance of multi-target tracking based on radar imagery to refine the maritime traffic situation awareness. In order to achieve this we simulate synthetic radar images and evaluate the tracking performance of different Bayesian multi-target trackers (MTTs), such as particle and JPDA filters. For the simulated tracks, the target state estimates in position, speed and course over ground will be compared to the reference data. The performance of the MTTs will be assessed via the OSPA metric by comparing the estimated multi-object state vector to the reference. This approach allows a fair performance analysis of different tracking algorithms based on radar images for a simulated maritime scenario.

show abstract

Section: Multi-target Trackersmentioning

confidence: 99%

Validation of Radar Image Tracking Algorithms with Simulated Data

Heymann¹,

Hoth²,

Banyś³

et al. 2017

TransNav

View full text Add to dashboard Cite

show abstract

“…However, the nonconvex energy formulation puts out of reach any possibility of global minimization. It is still possible to get approximate solutions using non-exact optimization techniques that do not require a specific energy formulation, as done in Multiple Hypothesis Tracking [17] using a breadth-first search with branch pruning or in Markov Chain Monte Carlo Data Association (MCMCDA) with MCMC sampling [20,29]. Despite the non-optimality of the found solution, these methods can fully exploit the use of more appropriate interaction and dynamic models and can therefore cope with more difficult tracking issues.…”

Section: Multi-frame Data Associationmentioning

confidence: 99%

“…Recently, many online or sliding window approaches have gained in performances by incorporating more complex appearance models [1,4,17]. These models, inspired by the recent improvements in Single Object Tracking (SOT), can be updated online to take into account changes in appearance or pose variations and help better distinguish targets, for more robust tracking results.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Improving Multi-frame Data Association with Sparse Representations for Robust Near-online Multi-object Tracking

Fagot-Bouquet

Audigier

Dhome

et al. 2016

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Multiple Object Tracking still remains a difficult problem due to appearance variations and occlusions of the targets or detection failures. Using sophisticated appearance models or performing data association over multiple frames are two common approaches that lead to gain in performances. Inspired by the success of sparse representations in Single Object Tracking, we propose to formulate the multi-frame data association step as an energy minimization problem, designing an energy that efficiently exploits sparse representations of all detections. Furthermore, we propose to use a structured sparsity-inducing norm to compute representations more suited to the tracking context. We perform extensive experiments to demonstrate the effectiveness of the proposed formulation, and evaluate our approach on two public authoritative benchmarks in order to compare it with several state-of-the-art methods.

show abstract

“…Track-by-detection algorithms use post-processed data, that is, the raw sensor measurements have had some kind of thresholding performed [47]. There is a vast amount of work currently being done on developing and evaluating new algorithms (since 2015) that fall into the former category [3,[48][49][50][51][52]. In track-by-detection approaches, there are often separate techniques for detection and data association.…”

Section: Current Trendsmentioning

confidence: 99%

Image-Based Multi-Target Tracking Through Multi-Bernoulli Filtering with Interactive Likelihoods

Hoak¹,

Medeiros²,

Povinelli³

2017

Preprint

View full text Add to dashboard Cite

Abstract:We develop an interactive likelihood (ILH) for sequential Monte-Carlo (SMC) methods for image-based multiple target tracking applications. The purpose of the ILH is to improve tracking accuracy by reducing the need for data association. In addition, we integrate a recently developed deep neural network for pedestrian detection along with the ILH with a multi-Bernoulli filter. We evaluate the performance of the multi-Bernoulli filter with the ILH and the pedestrian detector in a number of publicly available datasets (2003 PETS INMOVE, AFL, and TUD-Stadtmitte) using standard, well-known multi-target tracking metrics (OSPA and CLEAR MOT). In all datasets, the ILH term increases the tracking accuracy of the multi-Bernoulli filter.

show abstract

Multiple Hypothesis Tracking Revisited

Cited by 603 publications

References 33 publications

Validation of Radar Image Tracking Algorithms with Simulated Data

Validation of Radar Image Tracking Algorithms with Simulated Data

Improving Multi-frame Data Association with Sparse Representations for Robust Near-online Multi-object Tracking

Image-Based Multi-Target Tracking Through Multi-Bernoulli Filtering with Interactive Likelihoods

Contact Info

Product

Resources

About