The Bayesian inference has a two-step recursion structure, i.e., prediction and updating, which can be viewed as a dynamic reasoning process. Based on this elegant structure, various multi-target tracking (MTT) algorithms have been invented and successfully applied in many areas. On the other hand, Bayesian inference MTT algorithms are model-based methods that rely on models’ accuracy and first-order Markov assumption. In recent years, the MTT algorithms based on deep learning have received much attention due to their model-free property and the ability to learn from data, although they have issues such as over-fitting, generalization, etc. In this work, we propose a Transformer-based multi-target tracker whose architecture mimics the Bayesian inference, referred to as the Bayesian inference-based Transformer (BAIT) for MTT. To deal with the model mismatch issues, BAIT uses neural networks instead of the pre-assumed motion and observation models while retaining the excellent architecture of Bayesian inference. BAIT can recursively complete accurate predictions and updates via Transformer by refining the estimation of target states in a Bayesian inference-like manner. Thus, BAIT can be viewed as a combination of model-based and data-based methods. The simulation results show that, because of combining the advantages of Bayesian architecture with intelligent data association structure, BAIT is competitive in simple scenarios and achieves superior performance when the data association task becomes complicated.

Deep learning (DL) approaches, which do not rely on models and can learn complex relationships within data, garner increasing attention in the model-free multi-target tracking (MTT) domain. However, the study of applying the DL method to multi-sensor fusion-based MTT is relatively less. In this paper, we propose a Transformer-based distributed multi-sensor MTT approach, which adopts a hybrid fusion structure with both feature-level and decision-level fusion. First, for each local sensor, the high-dimensional feature information is extracted from the measurements based on a Transformer-based tracking module, which enables continuous tracking of multiple targets and provides the predicted target states and corresponding uncertainties. Then, the outputs of local sensors are fused using the covariance interception (CI) fusion rule. Finally, to further improve the fusion performance, the decision-level information is fed into a fusion decoder with the feature-level information to obtain the predicted target state and uncertainties after deep fusion. In this way, we realize a deep utilization of different sensors’ information and achieve a feature-level decision-level hybrid multi-sensor fusion, namely, Transformer-based multi-sensor hybrid fusion (TMSHF). Simulation results show that the proposed fusion method outperforms the CI algorithm in various tracking scenarios.

The swarm targets have shown great potential for both military and civilian applications, driving a high demand for reliable trajectory generation and accurate activity recognition. In this paper, we propose a trajectory generation method and establish an end-to-end deep learning model for dynamic continuous activity recognition of swarm targets. First, we devise an activity transition model of the drone swarm based on a continuous-time Markov chain (CTMC). Subsequently, the minimum snap trajectory generation algorithm is employed to generate the trajectories. After that, to recognize the dynamic continuous activity of targets, we develop an end-to-end neural network model to extract spatial and temporal features for swarm targets detected by radar across multiple frames. Finally, we demonstrate the effectiveness and robustness of our proposed method through simulation results.

Target detection in road clutter environment is a challenge for automotive radar. The performance of model-based methods degrades when the prior model is mismatched or the target energy is overwhelmed by the clutter. In contrast, deep learning methods can nonlinearly fit clutter distributions and extract deep features to identify targets from clutter backgrounds. Considering that the spatial-temporal feature in multi-frame data helps distinguish targets from clutter, we use the multi-frame data for detection. This paper proposes a multi-frame detection network for radar moving targets in clutter environment. First, we use transformer as the backbone to fit the large-scale clutter background by extracting the global spatio-temporal feature. Second, we proposed a multi-frame detection head to predict multi-frame bounding boxes in parallel by utilizing the spatio-temporal feature. Third, we proposed a segmentation-assisted refinement module to refine the objectness of bounding boxes, thus further suppressing the false alarms caused by clutter. Through experiments on simulation and measured datasets, the proposed method effectively reduces false alarms while maintaining a high detection probability. In addition, compared with the segmentation-based method, our method distinguishes adjacent targets more robustly.

Bit-Torrent is a well-known P2P application used to exchange digital content with high scalability. However, many users do not get authorization but transmit files through Bit-Torrent illegally. In this paper, we propose a DRM system that is suitable for the Bit-Torrent environment to protect copyrights, and also improve Chen et al.'s scheme in computational efficiency and security aspects.

Recently, Tsai presented an efficient algorithm which uses the ratio value of the hue over the intensity to construct the ratio map for detecting shadows of color aerial images. Instead of only using the global thresholding process in Tsai's algorithm, this paper presents a novel successive thresholding scheme (STS) to detect shadows more accurately. In our proposed STS, the modified ratio map, which is obtained by applying the exponential function to the ratio map proposed by Tsai, is presented to stretch the gap between the ratio values of shadow and nonshadow pixels. By performing the global thresholding process on the modified ratio map, a coarse-shadow map is constructed to classify the input color aerial image into the candidate shadow pixels and the nonshadow pixels. In order to detect the true shadow pixels from the candidate shadow pixels, the connected component process is first applied to the candidate shadow pixels for grouping the candidate shadow regions. For each candidate shadow region, the local thresholding process is performed iteratively to extract the true shadow pixels from the candidate shadow region. Finally, for the remaining candidate shadow regions, a fine-shadow determination process is applied to identify whether each remaining candidate shadow pixel is the true shadow pixel or not. Under six testing images, experimental results show that, for the first three testing images, both Tsai's and our proposed algorithms have better detection performance than that of the algorithm of Huang et al., and the shadow detection accuracy of our proposed STS-based algorithm is comparable to Tsai's algorithm. For the other three testing images, which contain some low brightness objects, our proposed algorithm has better shadow detection accuracy when compared with the previous two shadow detection algorithms proposed by Huang et al. and Tsai.