Aiming at the problems such as small object size, arbitrary object direction and complex background of remote sensing images, on the basis of YOLOv5 (You Only Look Once version 5) algorithm, an algorithm involved with geometric adaptation and global perception was proposed. Firstly, deformable convolutions and adaptive spatial attention modules were stacked alternately in series through dense connections. As a result, a Dense Context-Aware Module (DenseCAM) which can model local geometric features was constructed on the basis of taking full advantage of different levels of semantic and location information. Secondly, by introducing Transformer in the end of the backbone network, the global perception ability of the model was enhanced at a low cost and the relationships between objects and scenario content were modeled. On UCAS-AOD and RSOD datasets, compared with YOLOv5s6 algorithm, the proposed algorithm has the mean Average Precision (mAP) increased by 1.8 percentage points and 1.5 percentage points, respectively. Experimental results show that the proposed algorithm can effectively improve the precision of object detection in remote sensing images.
In the field of deep learning, a large number of correctly labeled samples are essential for model training. However, in practical applications, labeling data requires high labeling cost. At the same time, the quality of labeled samples is affected by subjective factors or tool and technology of manual labeling, which inevitably introduces label noise in the annotation process. Therefore, existing training data available for practical applications is subject to a certain amount of label noise. How to effectively train training data with label noise has become a research hotspot. Aiming at label noise learning algorithms based on deep learning, firstly, the source, classification and impact of label noise learning strategies were elaborated; secondly, four label noise learning strategies based on data, loss function, model and training method were analyzed according to different elements of machine learning; then, a basic framework for learning label noise in various application scenarios was provided; finally, some optimization ideas were given, and challenges and future development directions of label noise learning algorithms were proposed.
With the development and maturity of deep learning, the quality of neural machine translation has increased, yet it is still not perfect and requires human post-editing to achieve acceptable translation results. Interactive Machine Translation (IMT) is an alternative to this serial work, that is performing human interaction during the translation process, where the user verifies the candidate translations produced by the translation system and, if necessary, provides new input, and the system generates new candidate translations based on the current feedback of users, this process repeats until a satisfactory output is produced. Firstly, the basic concept and the current research progresses of IMT were introduced. Then, some common methods and state-of-the-art works were suggested in classification, while the background and innovation of each work were briefly described. Finally, the development trends and research difficulties of IMT were discussed.
Symbolic music generation is still an unsolved problem in the field of artificial intelligence and faces many challenges. It has been found that the existing methods for generating polyphonic music fail to meet the marke requirements in terms of melody, rhythm and harmony, and most of the generated music does not conform to basic music theory knowledge. In order to solve the above problems, a new Transformer-based multi-track music Generative Adversarial Network (Transformer-GAN) was proposed to generate music with high musicality under the guidance of music rules. Firstly, the decoding part of Transformer and the Cross-Track Transformer (CT-Transformer) adapted on the basis of Transformer were used to learn the information within a single track and between multiple tracks respectively. Then, a combination of music rules and cross-entropy loss was employed to guide the training of the generative network, and the well-designed objective loss function was optimized while training the discriminative network. Finally, multi-track music works with melody, rhythm and harmony were generated. Experimental results show that compared with other multi-instrument music generation models, for piano, guitar and bass tracks, Transformer-GAN improves Prediction Accuracy (PA) by a minimum of 12%, 11% and 22%, improves Sequence Similarity (SS) by a minimum of 13%, 6% and 10%, and improves the rest index by a minimum of 8%, 4% and 17%. It can be seen that Transformer -GAN can effectively improve the indicators including PA and SS of music after adding CT-Transformer and music rule reward module, which leads to a relatively high overall improvement of the generated music.
Aiming at the disadvantages of MD5 software such as large occupancy of resources and poor security, a hardware acceleration model of MD5 was put forward based on NetMagic platform, and then the non-streamlined and streamlined hardware acceleration models were verified and analyzed based on ModelSim and NetMagic platforms. Compared to non-streamlined acceleration model, the streamlined acceleration model can improve the operation efficiency of MD5 for five times. Its achievement can be applied to hardware encryption engine such as the network processor, which can effectively improve the security and processing efficiency of hardware equipment including the network processor.
To deal with the energy-efficient routing selection problem of the Wireless Sensor Network (WSN), an Energy-Efficient routing algorithm with Location information and Double cluster heads based on Hybrid Energy-Efficient Distributed clustering (HEED-EELD) was proposed. Assuming that all the network nodes had location awareness capabilities, the network was divided into different hierarchies according to the best single-hop distance, so the nodes determined their hierarchies based on their locations. Double cluster heads were selected to share a single cluster head's work and to balance the energy consumption. In the inter-cluster multi-hop routing, the cluster head selected the optimal route based on location, distance and cost function about residual energy. Matlab simulation results show that, compared with Low Energy Adaptive Clustering Hierarchy (LEACH) algorithm, HEED algorithm, HEED-EELD has obvious advantages in terms of network lifetime, energy efficiency and energy balancing.
To further reduce the great computational complexity for High Efficiency Video Coding (HEVC) intra prediction, a novel algorithm was proposed in this paper. First, in Coding Unit (CU) level, the minimum Sum of Absolute Transformed Difference (SATD) of current CU was used to decide an early termination for the split of this CU at each depth level: if the minimum SATD of this CU is smaller than the given threshold value. Meanwhile, based on statistical analysis, the probabilities of each candidate prediction modes being optimal mode were used to further reduce the number of candidate modes which have almost no chance to be selected as the best mode. The experimental results show that, the proposed algorithm can save an average of 30.5% of the encoding time with negligible loss of coding efficiency (only 0.02dB Y-PSNR(Y-Peak Signal-to-Noise Ratio) loss) compared with the reference model HM10.1. Besides, the proposed algorithm is easy to provide software and hardware implementations, and it is also easy to be combined with other methods to further reduce the great computational complexity for HEVC intra coding.
To improve the accuracy of traditional method for customer segmentation, the Large Average Submatrix (LAS) biclustering algorithm was used, which performed clusting on customer samples and consumer attributes simultaneously to identify the upscale and high-value customers. By introducing a new value yardstick and a novel index named PA, the LAS biclustering algorithm was compared with K-means clustering algorithm based on a simulation experiment on consumption data of a telecom corporation. The experimental result shows that the LAS biclustering algorithm finds more groups of high-value customers and obtains more accurate clusters. Therefore, it is more suitable for recognition and segmentation of high-value customers.