Online Learning Sample Filtering for Object Tracking

Despite many years of research, object tracking remains a challenging problem, not only because of the variety of object appearances, but also because of the complexity of surrounding environments. In this research, we present an algorithm for single object tracking using a particle filter framework and color histograms. Particle filters are iterative algorithms that perform predictions in each iteration using particles, which are samples drawn from a statistical distribution. Color histograms are embedded in these particles, and the distances between histograms are used to measure likelihood between targets and observations. One downside of color histograms is that they ignore spatial information, which may produce tracking failure when objects appear that are similar in color. To overcome this disadvantage, we propose a saliency-based weighting scheme for histogram calculation. Given an image region, first its saliency map is generated. Next, its histogram is calculated based on the generated saliency map. Pixels located in salient regions have higher weights than those in others, which helps preserve the spatial information. Experimental results showed the efficiency of the proposed appearance model in object tracking under various conditions.

An adaptive template matching-based single object tracking algorithm with parallel acceleration

Despite considerable similarities between multiple object tracking (MOT) and single object tracking (SOT) tasks, modern MOT methods have not benefited from the development of SOT ones to achieve satisfactory performance. The major reason for this situation is that it is inappropriate and inefficient to apply multiple SOT models directly to the MOT task, although advanced SOT methods are of the strong discriminative power and can run at fast speeds.In this paper, we propose a novel and end-to-end trainable MOT architecture that extends CenterNet by adding an SOT branch for tracking objects in parallel with the existing branch for object detection, allowing the MOT task to benefit from the strong discriminative power of SOT methods in an effective and efficient way. Unlike most existing SOT methods which learn to distinguish the target object from its local backgrounds, the added SOT branch trains a separate SOT model per target online to distinguish the target from its surrounding targets, assigning SOT models the novel discrimination. Moreover, similar to the detection branch, the SOT branch treats objects as points, making its online learning efficient even if multiple targets are processed simultaneously. Without tricks, the proposed tracker achieves MOTAs of 0.710 and 0.686, IDF1s of 0.719 and 0.714, on MOT17 and MOT20 benchmarks, respectively, while running at 16 FPS on MOT17.

Visual object tracking has become one of the most active research topics in computer vision, which has been growing in commercial development as well as academic research. Many visual trackers have been proposed in the last two decades. Recent studies of computer vision for dynamic scenes include motion detection, object classification, environment modeling, tracking of moving objects, understanding of object behaviors, object identification, and data fusion from multiple sensors. This paper provides an in-depth overview of recent object tracking research. Object tracking tasks in realistic scenario often face challenging problems such as camera motion, occlusion, illumination effect, clutter, and similar appearance. A variety of tracker techniques have been published, which combine multiple techniques to solve multiple visual tracking sub-problems. This paper also reviews the latest research trend in object tracking based on convolutional neural networks, which is receiving growing attention. Finally, the paper discusses the future challenges and research directions for the object tracking problems that still need extensive studies in coming years.

Object detection and tracking under complex environment are challenging because of the disturbances induced by background clutter, illumination changes, occlusions and other factors. The bulk of traditional algorithms basically rely on hand‐crafted features, which are not sufficiently robust to a complex environment. Moreover, the processes of detection and tracking are separated, which leads to the overall efficiency not high. In this study, a novel local probability model (LPM)‐based mean shift (MS) algorithm is proposed to integrate object detection and tracking. The main contributions include: (i) a new framework based on the combination of LPM and MS is established for the integration of object tracking and detection. (ii) For object detection, the training and prediction of LPM are built by stacked denoising autoencoders based deep learning. (iii) For object tracking, an MS tracking algorithm leveraging LPM is modified to improve the tracking efficiency under a complex environment. Experimental results demonstrate that the proposed method is superior to the colour histograms based MS and histograms of oriented gradients based MS in terms of robustness and tracking accuracy.

Most conventional object tracking algorithms are implemented on general-purpose processors in software due to its great flexibility. However, the real-time performance is hard to achieve due to the inherent characteristics of the sequential processing of these processors. To tackle this issue, a reconfigurable system-on-chip (rSoC) platform with microprocessors and FPGAs is applied in this paper. To simplify the hardware/software interface, a Belief–Desire–Intention (BDI)-based multi-agent architecture is proposed as the unified framework. Then an agent-based task graph and two heuristic partitioning methods are proposed to partition the hardware and software on an rSoC platform. Compared to the module-based architecture, this BDI-based multi-agent architecture provides more efficiency, flexibility, autonomy, and scalability for the real-time tracking systems. A particle swarm optimization (PSO)-based object detection and tracking algorithm is applied to evaluate the proposed architecture. Extensive experimental results of object tracking demonstrate that the proposed architecture is efficient and highly robust with real-time performance.

The demand for intelligent robots is increasing rapidly with the advancement in technology. This paper presents an end-to-end approach for a person following robot. It would be a great help if robots could behave as a helping hand for humans in critical environments like industries, markets as well as in hospitals. Because of balanced accuracy and speed, Siamese trackers got great attention. Here we are using a single object tracker for a person following and Robot Operating System (ROS) for communication between various nodes. The Robot is equipped with a Microsoft Kinect sensor which will provide RGBD (Red Green Blue image and Depth image) image. Initially, this system utilizes a deep learning algorithm for single object tracking to track a person. Further, the system can calculate the centroid of the target person as well as the depth information. Finally, this data is given to the navigation module of the robot. This will control the robot's motion

Moving object detection and tracking in video sequences is a task that emerges in various research fields of video processing, including video analysis and understanding, object-based coding and many related applications, such as content-based retrieval, remote surveillance and object recognition. This work revisits one of the most popular deformable templates for shape modeling and object tracking, the Snakes, and proposes a modified snake model and a probabilistic utilization of it for object tracking. Special attention has been drawn to complex natural (indoor and outdoor) sequences, where temporal clutter, abrupt motion and external lighting changes are crucial for the accuracy of the results, also focusing on the ability of the proposed approach to handle specific HCI problems, such as face and facial feature tracking. A variety of image sequences are used to illustrate the method's capability, providing theoretical explanation as well as experimental verification in specific tracking problems.

For underwater videos, the performance of object tracking is greatly affected by illumination changes, background disturbances and occlusion. Hence, there is a need to have a robust function that computes image similarity, to accurately track the moving object. In this work, a hybrid model that incorporates the Kalman Filter, a Siamese neural network and a miniature neural network has been developed for object tracking. It was observed that the usage of the Siamese network to compute image similarity significantly improved the robustness of the tracker. Although the model was developed for underwater videos, it was found that it performs well for both underwater and human surveillance videos. A metric has been defined for analyzing detections-to-tracks mapping accuracy. Tracking results have been analyzed using Multiple Object Tracking Accuracy (MOTA) and Multiple Object Tracking Precision (MOTP)metrics.

In this paper we address the basic limitation of SiamMask - the state of the art single object tracking and segmentation algorithm. SiamMask requires semi-supervision in that it needs a bounding box to be drawn manually around the object that has to be tracked. This is however not always possible or feasible, and slows down the pipeline even in the best case. We overcome this limitation by using state-of-the-art object detection algorithms: Detectron2 and YOLO to automatically detect the object and then track using SiamMask. The main purpose of this study is to devise an efficient technique for an end-to-end object detection and tracking, which can then be used in other applications like self-driving cars, etc. We compared different approaches using current state-of-the-art tools for time and detection efficiency. One of the secondary aim was to test how the two approaches perform on different types of datasets. We note that YOLO gives better and more meaningful detection of objects in the scene. However, Detectron2 gives a higher detection speed than YOLO, making the overall detection and tracking process faster.

A review of object tracking methods: From general field to autonomous vehicles

In recent years, the introduction of Siamese network has brought new vitality to the object tracking community. However, high-performance Siamese trackers cannot run at a real-time speed on mobile devices due to their complex and huge model. Knowledge distillation is a common and effective model compression method, but it is difficult to be applied to the challenging task like object tracking. We find out the fundamental cause is that the imbalance between the foreground and background in the object tracking task, which aggravates the problem of insufficient feature extraction ability of small backbone. Therefore, we propose the attention mask distillation (AMD) to help the student tracker focus on the foreground area faster and more accurately. The attention mask can be easily obtained from the feature maps and brings fine-granularity to the traditional binary mask. The experimental results on OTB 100 and VOT20 18 show that our method enables the student tracker perform as well as the teacher tracker. At the same time, it's able to run on the CPU at a hyper-real-time of 66 fps and achieves nearly 9 times model compression ratio. Such low computational and storage costs make it possible to deploy high-performance trackers on resource-constrained platforms.

In this paper, a modified particle swarm optimization (PSO) approach, particle swarm optimization with<i> ε- greedy</i> exploration εPSO), is used to tackle the object tracking. In the modified εPSO algorithm, the cooperative learning mechanism among individuals has been introduced, namely, particles not only adjust its own flying speed according to itself and the best individual of the swarm but also learn from other best individuals according to certain probability. This kind of biologically-inspired mutual-learning behavior can help to find the global optimum solution with better convergence speed and accuracy. The εPSO algorithm has been tested on benchmark function and demonstrated its effectiveness in high-dimension multi-modal optimization. In addition to the standard benchmark study, we also combined our new εPSO approach with the traditional particle filter (PF) algorithm on the object tracking task, such as car tracking in complex environment. Comparative studies between our εPSO combined PF algorithm with those of existing techniques, such as the particle filter (PF) and classic PSO combined PF will be used to verify and validate the performance of our approach.

As an effective feature extractor, Vision Transformer (ViT) has been widely applied to both image classification and object tracking tasks. In this paper, we revisit and enhance the classic Data-efficient image Transformer (DeiT) for these two tasks. The DeiT is optimized step-by-step across different modules, including its patch stem, position embedding, and the development of efficient linear attention mechanisms. To address the performance degradation of linear attention, we propose Shift Expansion Linear Attention (SELA) which generates new heads with rich feature diversity through a simple but efficient cyclic shift operation. Additionally, SELA similarity minimization is added to cross-entropy loss to further enhance feature diversity. Based on these improvements, we develop SELA-ViT for image classification and further build SELA-Track for object tracking. With comparable model size and speed, SELA-ViT-T achieves a +4.8% improvement in Top-1 accuracy over DeiT-T on ImageNet-1K and establishes a new state-of-the-art performance among linear attention methods. Furthermore, we validate SELA-ViT on five small datasets. On four benchmark object tracking datasets, SELA-Track exhibits improved tracking performance. The code and models are available at: https://github.com/saizhou777/SELA-ViT.

Segmenting and tracking of objects in video is of great importance for video-based encoding, surveillance, and retrieval. However, the inherent difficulty of object segmentation and tracking is to distinguish changes in the displacement of objects from disturbing effects such as noise and illumination changes. Therefore, in this paper, we formulate a color-based deformable model which is robust against noisy data and changing illumination. Computational methods are presented to measure color constant gradients. Further, a model is given to estimate the amount of sensor noise through these color constant gradients. The obtained uncertainty is subsequently used as a weighting term in the deformation process. Experiments are conducted on image sequences recorded from three-dimensional scenes. From the experimental results, it is shown that the proposed color constant deformable method successfully finds object contours robust against illumination, and noisy, but homogeneous regions.