Tracking Noisy Targets: A Review of Recent Object Tracking Approaches

Fiaz, Mustansar; Mahmood, Arif; Jung, Soon Ki

doi:10.48550/arxiv.1802.03098

Cited by 10 publications

(11 citation statements)

References 164 publications

(287 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We, therefore, designed our algorithm to detect, track and re-identify the object of interest across several non-overlapping cameras using the multi-object tracking process. We implemented the proposed algorithm using the dataset that contains different poses of persons [ 22 ] and different illumination conditions. The algorithm is divided into two modules, namely, detection and tracking.…”

Section: Proposed Hcnn For Real-time Motmentioning

confidence: 99%

Enhancing Detection Quality Rate with a Combined HOG and CNN for Real-Time Multiple Object Tracking across Non-Overlapping Multiple Cameras

Kalake

Dong

Wan

et al. 2022

Sensors

View full text Add to dashboard Cite

Multi-object tracking in video surveillance is subjected to illumination variation, blurring, motion, and similarity variations during the identification process in real-world practice. The previously proposed applications have difficulties in learning the appearances and differentiating the objects from sundry detections. They mostly rely heavily on local features and tend to lose vital global structured features such as contour features. This contributes to their inability to accurately detect, classify or distinguish the fooling images. In this paper, we propose a paradigm aimed at eliminating these tracking difficulties by enhancing the detection quality rate through the combination of a convolutional neural network (CNN) and a histogram of oriented gradient (HOG) descriptor. We trained the algorithm with an input of 120 × 32 images size and cleaned and converted them into binary for reducing the numbers of false positives. In testing, we eliminated the background on frames size and applied morphological operations and Laplacian of Gaussian model (LOG) mixture after blobs. The images further underwent feature extraction and computation with the HOG descriptor to simplify the structural information of the objects in the captured video images. We stored the appearance features in an array and passed them into the network (CNN) for further processing. We have applied and evaluated our algorithm for real-time multiple object tracking on various city streets using EPFL multi-camera pedestrian datasets. The experimental results illustrate that our proposed technique improves the detection rate and data associations. Our algorithm outperformed the online state-of-the-art approach by recording the highest in precisions and specificity rates.

show abstract

Section: Proposed Hcnn For Real-time Motmentioning

confidence: 99%

Enhancing Detection Quality Rate with a Combined HOG and CNN for Real-Time Multiple Object Tracking across Non-Overlapping Multiple Cameras

Kalake

Dong

Wan

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

“…Fiaz et al [6] noted that deep learning-based tracking algorithms rely on appearance: they use the correlation between appearance-based feature maps of successive frames to track target objects. More recent methods have shown incremental improvements by using this same approach on object-proposals from instance segmentation models applied to individual frames [2].…”

Section: Benchmark Models: 3d-cnnsmentioning

confidence: 99%

The Challenge of Appearance-Free Object Tracking with Feedforward Neural Networks

Malik¹,

Linsley²,

Serre³

et al. 2021

Preprint

View full text Add to dashboard Cite

Nearly all models for object tracking with artificial neural networks depend on appearance features extracted from a "backbone" architecture, designed for object recognition. Indeed, significant progress on object tracking has been spurred by introducing backbones that are better able to discriminate objects by their appearance. However, extensive neurophysiology and psychophysics evidence suggests that biological visual systems track objects using both appearance and motion features [1]. Here, we introduce Pathtracker, a visual challenge inspired by cognitive psychology, which tests the ability of observers to learn to track objects solely by their motion. We find that standard 3Dconvolutional deep network models struggle to solve this task when clutter is introduced into the generated scenes, or when objects travel long distances. This challenge reveals that tracing the path of object motion is a blind spot of feedforward neural networks. We expect that strategies for appearance-free object tracking from biological vision can inspire solutions these failures of deep neural networks.

show abstract

“…Recently, deep-learning based object trackers have shown promising results for many practical problems. In particular, DeepSort [6] uses a deep-learning based object detector combined with the Kalman filter [7] to carry out multi-object tracking, which is the task of tracking a set of objects in a sequence of frames [8]. The object detector generates detection results as bounding boxes, which are used to initialize the trackers.…”

Section: Related Workmentioning

confidence: 99%

Turkey Behavior Identification System with a GUI Using Deep Learning and Video Analytics

Ju¹,

Mahapatra²,

Erasmus³

et al. 2021

View full text Add to dashboard Cite

show abstract

Tracking Noisy Targets: A Review of Recent Object Tracking Approaches

Cited by 10 publications

References 164 publications

Enhancing Detection Quality Rate with a Combined HOG and CNN for Real-Time Multiple Object Tracking across Non-Overlapping Multiple Cameras

Enhancing Detection Quality Rate with a Combined HOG and CNN for Real-Time Multiple Object Tracking across Non-Overlapping Multiple Cameras

The Challenge of Appearance-Free Object Tracking with Feedforward Neural Networks

Turkey Behavior Identification System with a GUI Using Deep Learning and Video Analytics

Contact Info

Product

Resources

About