QEST: Quantized and Efficient Scene Text Detector using Deep Learning

Small target detection algorithm based on multi-branch stacking and new sampling transition module

Aiming at the problem that the SSD algorithm does not fully extract the feature information contained in each feature layer, as well as the feature information is easily lost during the sampling process, which makes the feature expression ineffective and leads to insufficient performance in small target detection. In this paper, AMT-SSD is proposed, a small target detection algorithm that incorporates the multi-branch stacking and new sampling transition module of the attention mechanism. In this algorithm, the composite attention mechanism is utilized to improve the correlation of features of the samples to be detected in terms of spatial and channels, and the efficiency of the algorithm; secondly, multi-branch stacking module is used to extract multi-size features for each feature layer, and different sizes of convolution kernels are utilized in parallel to fully extract their features and improve the expression of features; meanwhile, during the sampling process, the problem of missing features is solved by applying inverse subpixel convolution in the new sampling transition module. Experimentally, the AMT-SSD algorithm achieves 84.6% and 53.4% mAP metrics on the PASCAL VOC dataset and MS COCO dataset, respectively. This indicates that the AMT-SSD algorithm can effectively extract feature information that is beneficial to detection samples, and also performs well in reducing feature loss, which is effective for the algorithm to improve the algorithm on small targets.

Novel Deep-Learning Modulation Recognition Algorithm Using 2D Histograms over Wireless Communications Channels

Modulation recognition (MR) has become an essential topic in today's wireless communications systems. Recently, convolutional neural networks (CNNs) have been employed as a potent tool for MR because of their ability to minimize the feature's susceptibility to its surroundings and reduce the need for human feature extraction and evaluation. In particular, these investigations rely on the unrealistic assumption that the channel coefficient is typically one. This motivates us to overcome the previous constraint by providing a novel MR suited to fading wireless channels. This paper proposes a novel MR algorithm that is capable of recognizing a broad variety of modulation types, including M-ary QAM and M-ary PSK, without enforcing any restrictions on the modulation size, M. The analysis has shown that each modulation choice has a distinct two-dimensional in-phase quadrature histogram. This property is beneficially utilized to design a convolutional neural-network-based MR algorithm. When compared to the existing techniques, Monte Carlo simulations demonstrated the success of the proposed design.

Neural Graph Similarity Computation with Contrastive Learning

Computing the similarity between graphs is a longstanding and challenging problem with many real-world applications. Recent years have witnessed a rapid increase in neural-network-based methods, which project graphs into embedding space and devise end-to-end frameworks to learn to estimate graph similarity. Nevertheless, these solutions usually design complicated networks to capture the fine-grained interactions between graphs, and hence have low efficiency. Additionally, they rely on labeled data for training the neural networks and overlook the useful information hidden in the graphs themselves. To address the aforementioned issues, in this work, we put forward a contrastive neural graph similarity learning framework, Conga. Specifically, we utilize vanilla graph convolutional networks to generate the graph representations and capture the cross-graph interactions via a simple multilayer perceptron. We further devise an unsupervised contrastive loss to discriminate the graph embeddings and guide the training process by learning more expressive entity representations. Extensive experiment results on public datasets validate that our proposal has more robust performance and higher efficiency compared with state-of-the-art methods.