Search Results (7)

Understanding how piglets move around sows during posture changes is crucial for their safety and healthy growth. Automated monitoring can reduce farm labor and help prevent accidents like piglet crushing. Current methods (called Joint Detection-and-Tracking-based, abbreviated as JDT-based) struggle with problems like misidentifying piglets or losing track of them due to crowding, occlusion, and shape changes. To solve this, we developed MSHMTracker, a smarter tracking system that introduces a motion-status hierarchical architecture to significantly improve tracking performance by adapting to piglets’ motion statuses. In MSHMTracker, a score- and time-driven hierarchical matching mechanism (STHM) was used to establish the spatio-temporal association by the motion status, helping maintain accurate tracking even in challenging conditions. Finally, piglet group aggregation or dispersion behaviors in response to sow posture changes were identified based on the tracked trajectory information. Tested on 100 videos (30,000+ images), our method achieved 93.8% tracking accuracy (MOTA) and 92.9% identity consistency (IDF1). It outperformed six popular tracking systems (e.g., DeepSort, FairMot). The mean accuracy of behavior recognition was 87.5%. In addition, the correlations (0.6 and 0.82) between piglet stress responses and sow posture changes were explored. This research showed that piglet movements are closely related to sow behavior, offering insights into sow–piglet relationships. This work has the potential to reduce farmers’ labor and improve the productivity of animal husbandry. Full article

Airborne radar-based ground moving target tracking faces challenges such as low detection rates and high clutter density. While lowering the detection threshold can improve detection performance, it introduces significant false alarms, thereby degrading tracking performance. To address these challenges, this paper proposes a novel multi-information assisted Joint Detection and Tracking (JDT) framework for ground moving targets. This study enhances detection and tracking performance by integrating multi-source information, specifically echo information, road network data, and velocity limits, enabling bidirectional data exchange between the detector and tracker for multiple ground targets. An adaptive threshold detector is developed by incorporating a priori information and tracker feedback. Additionally, we innovatively propose an improved Variable Structure Interacting Multiple Model (VS-IMM) filter that leverages road network constraints and detector outputs for tracking, featuring an enhanced model probability calculation to significantly reduce computational time. Simulation results demonstrate that the proposed method significantly improves data association accuracy and tracking precision. Full article

Multi-object tracking (MOT) in satellite videos is a challenging task due to the small size and blurry features of objects, which often lead to intermittent detection and tracking instability. Many existing object detection and tracking models often struggle with these issues, as they are not designed to effectively handle the unique characteristics of satellite videos. To address these challenges, we propose LocaLock, a joint detection and tracking framework for MOT that incorporates feature matching concepts from single object tracking (SOT) to enhance tracking stability and reduce intermittent tracking results. Specifically, LocaLock utilizes an anchor-free detection backbone for efficiency and employs a local cost volume (LCV) module to perform precise feature matching in the local area. This provides valuable object priors to the detection head, enabling the model to “lock” onto objects with greater accuracy and mitigate the instability associated with small object detection. Additionally, the local computation within the LCV module ensures low computational complexity and memory usage. Furthermore, LocaLock incorporates a novel motion flow (MoF) module to accumulate and exploit temporal information, further enhancing feature robustness and consistency across frames. Rigorous evaluations on the VISO dataset demonstrate the superior performance of LocaLock, surpassing existing methods in tracking accuracy and precision within the demanding satellite video analysis domain. Notably, LocaLock achieved state-of-the-art performance on the VISO benchmark, achieving a multi-object tracking accuracy (MOTA) of 62.6 while ensuring fast running speed. Full article

In this paper, we focus on the multi-target tracking (MOT) task in satellite videos. To achieve efficient and accurate tracking, we propose a transformer-distillation-based end-to-end joint detection and tracking (JDT) method. Specifically, (1) considering that targets in satellite videos usually have small scales and are shot from a bird’s-eye view, we propose a pixel-wise transformer-based feature distillation module through which useful object representations are learned via pixel-wise distillation using a strong teacher detection network; (2) targets in satellite videos, such as airplanes, ships, and vehicles, usually have similar appearances, so we propose a temperature-controllable key feature learning objective function, and by highlighting the learning of similar features during distilling, the tracking accuracy for such objects can be further improved; (3) we propose a method that is based on an end-to-end network but simultaneously learns from a highly precise teacher network and tracking head during training so that the tracking accuracy of the end-to-end network can be improved via distillation without compromising efficiency. The experimental results on three recently released publicly available datasets demonstrated the superior performance of the proposed method for satellite videos. The proposed method achieved over 90% overall tracking performance on the AIR-MOT dataset. Full article

Multiple object tracking (MOT) plays an important role in intelligent video-processing tasks, which aims to detect and track all moving objects in a scene. Joint-detection-and-tracking (JDT) methods are thriving in MOT tasks, because they accomplish the detection and data association in a single stage. However, the slow training convergence and insufficient data association limit the performance of JDT methods. In this paper, the anchor-based query (ABQ) is proposed to improve the design of the JDT methods for faster training convergence. By augmenting the coordinates of the anchor boxes into the learnable queries of the decoder, the ABQ introduces explicit prior spatial knowledge into the queries to focus the query-to-feature learning of the JDT methods on the local region, which leads to faster training speed and better performance. Moreover, a new template matching (TM) module is designed for the JDT methods, which enables the JDT methods to associate the detection results and trajectories with historical features. Finally, a new transformer-based MOT method, ABQ-Track, is proposed. Extensive experiments verify the effectiveness of the two modules, and the ABQ-Track surpasses the performance of the baseline JDT methods, TransTrack. Specifically, the ABQ-Track only needs to train for 50 epochs to achieve convergence, while that for TransTrack is 150 epochs. Full article

Traditional target tracking is carried out based on the point measurements extracted from the radar resolution cells. This is not suitable for situations of low signal-to-noise ratio (SNR). In this paper, we aim to investigate the problem of the joint detection and tracking (JDT) of a target by directly using the received signals of passive multi-static radar without feeding the signals to matched filters. To this end, a novel likelihood function is proposed exploiting the statistical properties of coherent processing between the reference and surveillance signals. With such a likelihood function, the particle Bernoulli filter is employed to perform direct JDT (DJDT) of the target. A remarkable feature of the proposed method is that it is able to achieve satisfactory performance when the SNR of received signals is low. Furthermore, the proposed method cannot only achieve the existence and kinematic state of the target, but also the time-varying SNR of each receiver, which serves as an important input for sensor adjustment. The performance of the proposed method is verified via simulations. Full article

Multi-target tracking, a high-level vision job in computer vision, is crucial to understanding autonomous driving surroundings. Numerous top-notch multi-object tracking algorithms have evolved in recent years as a result of deep learning’s outstanding performance in the field of visual object tracking. There have been a number of evaluations on individual sub-problems, but none that cover the challenges, datasets, and algorithms associated with visual multi-object tracking in autonomous driving scenarios. In this research, we present an exhaustive study of algorithms in the field of visual multi-object tracking over the last ten years, based on a systematic review approach. The algorithm is broken down into three groups based on its structure: methods for tracking by detection (TBD), joint detection and tracking (JDT), and Transformer-based tracking. The research reveals that the TBD algorithm has a straightforward structure, however the correlation between its individual sub-modules is not very strong. To track multiple objects, the JDT technique combines multi-module joint learning with a deep network framework. Transformer-based algorithms have been explored over the past two years, and they have benefits in numerous assessment indicators, as well as tremendous research potential in the area of multi-object tracking. Theoretical support for algorithmic research in adjacent disciplines is provided by this paper. Additionally, the approach we discuss, which uses merely monocular cameras rather than sophisticated sensor fusion, is anticipated to pave the way for the quick creation of safe and affordable autonomous driving systems. Full article