There are many Dynamic Multiobjective Optimization Problems (DMOPs) in real life. For such problems, when the environment changes, Dynamic Multi-Objective Evolutionary Algorithm (DMOEA) is required to track the Pareto Front (PF) or Pareto Set (PS) quickly and accurately under the new environment. Aiming at the problem of poor performance of the existing algorithms on population prediction, a dynamic multi-objective optimization algorithm based on Weight Vector Clustering Prediction (WVCP) was proposed. Firstly, the uniform weight vectors were generated in the target space, and the individuals in the population were clustered. According to the clustering results, the distribution of the population was analyzed. Secondly, a time series was established for the center points of clustered individuals. For the same weight vector, the corresponding coping strategies were adopted to supplement individuals according to different clustering situations. If there were cluster centers at all adjacent moments, the difference model was used to predict individuals in the new environment. If there was no cluster center at a certain moment, the centroid of the cluster centers of adjacent weight vectors was used as the cluster center at that moment, and then the difference model was used to predict individuals. In this way, the problem of poor population distribution was solved effectively, and the accuracy of prediction was improved at the same time. Finally, the introduction of individual supplement strategy was beneficial to make full use of historical information. In order to verify the performance of the proposed algorithm, simulation comparison of this algorithm and four representative algorithms was carried out. Experimental results show that the proposed algorithm can solve DMOPs well.
To solve severe performance degradation problem of person re-identification task during cross-domain migration, a new cross-domain person re-identification method based on attention mechanism with learning intra-domain variance was proposed. Firstly, ResNet50 was used as the backbone network and some modifications were made to it, so that it was more suitable for person re-identification task. And Instance-Batch Normalization Network (IBN-Net) was introduced to improve the generalization ability of model. At the same time, for the purpose of learning more discriminative features, a region attention branch was added to the backbone network. For the training of source domain, it was treated as a classification task. Cross-entropy loss was utilized for supervised learning of source domain, and triplet loss was introduced to mine the details of source domain samples and improve the classification performance of source domain. For the training of target domain, intra-domain variance was considered to adapt the difference in data distribution between the source domain and the target domain. In the test phase, the output of ResNet50 pool-5 layer was used as image features, and Euclidean distance between query image and candidate image was calculated to measure the similarity of them. In the experiments on two large-scale public datasets of Market-1501 and DukeMTMC-reID, the Rank-1 accuracy of the proposed method is 80.1% and 67.7% respectively, and its mean Average Precision (mAP) is 49.5% and 44.2% respectively. Experimental results show that, the proposed method has better performance in improving generalization ability of model.
In order to solve the problem of huge labeling cost for person re-identification, a method of one-shot video-based person re-identification with multi-loss learning and joint metric was proposed. Aiming at the problem that the number of label samples is small and the model obtained is not robust enough, a Multi-Loss Learning (MLL) strategy was proposed. In each training process, different loss functions were used for different data to optimize and improve the discriminative ability of the model. Secondly, a Joint Distance Metric (JDM) was proposed for label estimation, which combined the sample distance and the nearest neighbor distance to further improve the accuracy of pseudo label prediction. JDM solved the problems of the low accuracy of label estimation for unlabeled data, and the instability in the training process caused by the unlabeled data not fully utilized. Experimental results show that compared with the one-shot progressive learning method PL (Progressive Learning), the rank-1 accuracy reaches 65.5% and 76.2% on MARS and DukeMTMC-VideoReID datasets when the ratio of pseudo label samples added per iteration is 0.10, with the improvement of the proposed method of 7.6 and 5.2 percentage points, respectively.
To address the performance degradation of segmentation-based text detection methods in aliasing text scenes, a Single Direction Projected Transformer (SDPT) was proposed for aliasing text detection. Firstly, multi-scale features were extracted and fused by using deep Residual Network (ResNet) and Feature Pyramid Network (FPN). Then, the feature map was projected into a vector sequence by using horizontal projection and was fed into the Transformer module to model, thereby mining the relationship between the lines of text. Finally, joint optimization was performed using multiple objectives. Extensive experiments were conducted on the synthetic dataset BDD-SynText and the real dataset RealText. The results show that the proposed SDPT achieves optimal effect for text detection with high aliasing level, and improves F1-Score (IoU75) by at least 21. 36 percentage points on BDD-SynText and 18.11 percentage points on RealText compared with the state-of-the-art text detection algorithms such as Progressive Scale Expansion Network (PSENet) under the same backbone network (ResNet50), verifying the important role of the proposed method for performance improvement in aliasing text detection.
Single Long Short-Term Memory (LSTM) network cannot effectively extract key information and cannot accurately fit data distribution in trajectory prediction. In order to solve the problems, a short-term trajectory prediction model of aircraft based on attention mechanism and Generative Adversarial Network (GAN) was proposed. Firstly, different weights were assigned to the trajectory by introducing attention mechanism, so that the influence of important features in the trajectory was able to be improved. Secondly, the trajectory sequence features were extracted by using LSTM, and the convergence net was used to gather all aircraft features within the time step. Finally, the characteristic of GAN optimizing continuously in adversarial game was used to optimize the model in order to improve the model accuracy. Compared with Social Generative Adversarial Network (SGAN), the proposed model has the Average Displacement Error (ADE), Final Displacement Error (FDE) and Maximum Displacement Error (MDE) reduced by 20.0%, 20.4% and 18.3% respectively on the dataset during climb phase. Experimental results show that the proposed model can predict future trajectories more accurately.
In order to reduce the complexity of signal reconstruction algorithm, and reconstruct the signal with unknown sparsity, a new algorithm named One Projection Subspace Pursuit (OPSP) was proposed. Firstly, the upper and lower bounds of the signal's sparsity were determined based on the restricted isometry property, and the signal's sparsity was set as their integer middle value. Secondly, under the frame of Subspace Pursuit (SP), the projection of the observation onto the support set in each iteration process was removed to decrease the computational complexity of the algorithm. Furthermore, the whole signal's reconstruction rate was used as the index of reconstruction performance. The simulation results show that the proposed algorithm can reconstruct the signals of unknown sparsity with less time and higher reconstruction rate compared with the traditional SP algorithm, and it is effective for signal reconstruction.
Next Generation Network (NGN) is an integrative network which uses different radio access technologies. In this converged network environment, vertical handoff between different wireless access technologies becomes an important research topic. However, most of vertical handoff algorithms do not think about the actual demands of network and the mobility of user, but taking network properties as the standards of judgment. In order to solve the problem above, a speed adaptive vertical handoff algorithm based on application requirements was proposed, which used the speed factor and network propertise matrix to compensate for the quality loss of wireless link caused by mobility, which adaptively adjusted the weights of network properties that the application needs and supported node to make effective decisions. This algorithm realized vertical handoff with adaptive speed which better served the application and . Simulation results show that the proposed algorithm can overcome the ping-pang effect effectively and it has higher packet throughput in comparison with the other vertical handoff algorithms.