No abstract

No abstract

In order to solve the problems of filtering divergence and low accuracy in Kalman filter (KF) applications in a high-speed unmanned aerial vehicle (UAV), this paper proposed a new method of integrated robust adaptive Kalman filter: strong adaptive Kalman filter (SAKF). The simulation of two high-dynamic conditions and a practical experiment were designed to verify the new multi-sensor data fusion algorithm. Then the performance of the Sage–Husa adaptive Kalman filter (SHAKF), strong tracking filter (STF), H∞ filter and SAKF were compared. The results of the simulation and practical experiments show that the SAKF can automatically select its filtering process under different conditions, according to an anomaly criterion. SAKF combines the advantages of SHAKF, H∞ filter and STF, and has the characteristics of high accuracy, robustness and good tracking skill. The research has proved that SAKF is more appropriate in high-speed UAV navigation than single filter algorithms.

No abstract

Due to the low costs of its storage and search, the cross-modal retrieval hashing method has received much research interest in the big data era. Due to the application of deep learning, the cross-modal representation capabilities have risen markedly. However, the existing deep hashing methods cannot consider multi-label semantic learning and cross-modal similarity learning simultaneously. That means potential semantic correlations among multimedia data are not fully excavated from multi-category labels, which also affects the original similarity preserving of cross-modal hash codes. To this end, this paper proposes deep multi-semantic fusion-based cross-modal hashing (DMSFH), which uses two deep neural networks to extract cross-modal features, and uses a multi-label semantic fusion method to improve cross-modal consistent semantic discrimination learning. Moreover, a graph regularization method is combined with inter-modal and intra-modal pairwise loss to preserve the nearest neighbor relationship between data in Hamming subspace. Thus, DMSFH not only retains semantic similarity between multi-modal data, but integrates multi-label information into modal learning as well. Extensive experimental results on two commonly used benchmark datasets show that our DMSFH is competitive with the state-of-the-art methods.