Multiple People Tracking by Lifted Multicut and Person Re-identification

Tang, Siyu; Andriluka, Mykhaylo; Andres, Bjoern; Schiele, Bernt

doi:10.1109/cvpr.2017.394

Cited by 511 publications

(368 citation statements)

References 26 publications

Supporting

Mentioning

357

Contrasting

Order By: Relevance

“…MOT with Deep Learning. Recently, deep learning has been applied to multi-object tracking [28,21,31,38,16,36,4,37]. The trend in this line is to learn deep representations [23,30,39,36], and then employ traditional assignment strategies such as bipartite matching [36], or linear assignment for optimization [38].…”

Section: Related Workmentioning

confidence: 99%

“…Leal-Taixet al [21] propose a Siamese CNN to estimate the similarity between targets and detections. Tang et al [37] go a step further by treating MOT as the person Re-ID problem and develop a Siamese ID-Net to compute association costs between detections. Sadeghian et al [31] exploit CNN and LSTM to build the affinity measures based on appearance, motion and interaction cues.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

End-to-End Learning Deep CRF Models for Multi-Object Tracking Deep CRF Models

Xiang

et al. 2021

IEEE Trans. Circuits Syst. Video Technol.

View full text Add to dashboard Cite

Existing deep multi-object tracking (MOT) approaches first learn a deep representation to describe target objects and then associate detection results by optimizing a linear assignment problem. Despite demonstrated successes, it is challenging to discriminate target objects under mutual occlusion or to reduce identity switches in crowded scenes. In this paper, we propose learning deep conditional random field (CRF) networks, aiming to model the assignment costs as unary potentials and the long-term dependencies among detection results as pairwise potentials. Specifically, we use a bidirectional long short-term memory (LSTM) network to encode the long-term dependencies. We pose the CRF inference as a recurrent neural network learning process using the standard gradient descent algorithm, where unary and pairwise potentials are jointly optimized in an end-to-end manner. Extensive experimental results on the challenging MOT datasets including MOT-2015 and MOT-2016, demonstrate that our approach achieves the state of the art performances in comparison with published works on both benchmarks.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

End-to-End Learning Deep CRF Models for Multi-Object Tracking Deep CRF Models

Xiang

et al. 2021

IEEE Trans. Circuits Syst. Video Technol.

View full text Add to dashboard Cite

show abstract

“…Most MOT methods belong to the first category and process the video in an offline way, where the data association is optimized over the whole video or a span of frames and requires future frames to determine objects' states in the current frame. Network flow-based MOT methods [25,26] are quite typical in this category, and they generally solve the MOT problem using minimumcost flow optimization. In [25], linking person hypotheses over time is formulated as a minimum cost lifted multicut problem.…”

Section: Data Associationmentioning

confidence: 99%

Instance-aware representation learning and association for online multi-person tracking

Wang

et al. 2019

Pattern Recognition

View full text Add to dashboard Cite

Multi-Person Tracking (MPT) is often addressed within the detection-to-association paradigm. In such approaches, human detections are first extracted in every frame and person trajectories are then recovered by a procedure of data association (usually offline). However, their performances usually degenerate in presence of detection errors, mutual interactions and occlusions. In this paper, we present a deep learning based MPT approach that learns instance-aware representations of tracked persons and robustly online infers states of the tracked persons. Specifically, we design a multibranch neural network (MBN), which predicts the classification confidences and locations of all targets by taking a batch of candidate regions as input. In our MBN architecture, each branch (instance-subnet) corresponds to an individual to be tracked and new branches can be dynamically created for handling newly appearing persons. Then based on the output of MBN, we construct a joint association matrix that represents meaningful states of tracked persons (e.g., being tracked or disappearing from the scene) and solve it by using the efficient Hungarian algorithm. Moreover, we allow the instance-subnets to be updated during tracking by online mining hard examples, accounting to person appearance variations over time. We comprehensively evaluate our framework on a popular MPT benchmark, demonstrating its excellent performance in comparison with recent online MPT methods.

show abstract

“…Since the quest for algorithms that enable cognitive abilities is an important part of machine learning, person reidentification (ReID) has become more attractive, where the model is requested to be capable of correctly matching images of pedestrians across videos captured from different cameras. This task has drawn increasing attention in many computer vision applications, such as surveillance [49], activity analysis [31,32] and people tracking [55,44]. It is also challenging because the images of pedestrians are captured from disjoint views, the lighting-conditions/personposes differ across cameras, and occlusions are frequent in real-world scenarios.…”

Section: Introductionmentioning

confidence: 99%

Mixed High-Order Attention Network for Person Re-Identification

Chen¹,

Deng

2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

360

188

View full text Add to dashboard Cite

Attention has become more attractive in person reidentification (ReID) as it is capable of biasing the allocation of available resources towards the most informative parts of an input signal. However, state-of-the-art works concentrate only on coarse or first-order attention design, e.g. spatial and channels attention, while rarely exploring higher-order attention mechanism. We take a step towards addressing this problem. In this paper, we first propose the High-Order Attention (HOA) module to model and utilize the complex and high-order statistics information in attention mechanism, so as to capture the subtle differences among pedestrians and to produce the discriminative attention proposals. Then, rethinking person ReID as a zero-shot learning problem, we propose the Mixed High-Order Attention Network (MHN) to further enhance the discrimination and richness of attention knowledge in an explicit manner. Extensive experiments have been conducted to validate the superiority of our MHN for person ReID over a wide variety of state-of-the-art methods on three large-scale datasets, including Market-1501, DukeMTMC-ReID and CUHK03-NP. Code is available at

show abstract

Multiple People Tracking by Lifted Multicut and Person Re-identification

Cited by 511 publications

References 26 publications

End-to-End Learning Deep CRF Models for Multi-Object Tracking Deep CRF Models

End-to-End Learning Deep CRF Models for Multi-Object Tracking Deep CRF Models

Instance-aware representation learning and association for online multi-person tracking

Mixed High-Order Attention Network for Person Re-Identification

Contact Info

Product

Resources

About