Most of the existing steganalysis methods for low-bit-rate compressed speech are specifically designed for a particular speech encoder or category of steganography methods, limiting their generalization capability. These methods require pre-selection of codewords affected by the specific steganographic process as input to the steganalysis models. In order to overcome this limitation and enhance the practicality of steganalysis algorithms, we propose a compressed S peech encoder and steganography A lgorithm independent steganalysis N etwork, named SANet . Irrespective of the specific steganography algorithm used, modifications to the codewords will impact the sequential correlation characteristics of uncompressed domain (time domain) speech. Additionally, the compressed speech streams from different coders are unified in the uncompressed domain format. Therefore, this article introduces an intermediate representation based on the uncompressed domain and develops a neural network that utilizes collaborative correlation features to extract steganography-sensitive characteristics from this representation. Experimental results demonstrate that our proposed method achieves state-of-the-art detection performance for various steganography algorithms under different speech encoders.

This article proposes a radiation-stealth integrated antenna structure based on an FSR (Frequency Selective Rasorber) structure. When the antenna is operational, this structure functions as a normal radiator; however, when it is inactive, the entire system becomes a broadband absorber. Simulation results indicate that the structure exhibits a VSWR < 3 bandwidth spanning from 9.69 GHz to 10.63 GHz, with boresight radiation gains exceeding 5.4 dBic. In non-operational state, the antenna structure achieves broadband absorption across the frequency range of 4.8 GHz to 15.48 GHz.

Avian metapneumovirus (aMPV) is a highly contagious pathogen that causes acute upper respiratory tract diseases in chickens and turkeys, resulting in serious economic losses. Subtype B aMPV has recently become the dominant epidemic strain in China. We developed an attenuated aMPV subtype B strain by serial passaging in Vero cells and evaluated its safety and efficacy as a vaccine candidate. The safety test showed that after the 30th passage, the LN16-A strain was fully attenuated, as clinical signs of infection and histological lesions were absent after inoculation. The LN16-A strain did not revert to a virulent strain after five serial passages in chickens. The genomic sequence of LN16-A differed from that of the parent wide-type LN16 (wtLN16) strain and had nine amino acid mutations. In chickens, a single immunization with LN16-A induced robust humoral and cellular immune responses, including the abundant production of neutralizing antibodies, CD4+ T lymphocytes, and the Th1 (IFN-γ) and Th2 (IL-4 and IL-6) cytokines. We also confirmed that LN16-A provided 100% protection against subtype B aMPV and significantly reduced viral shedding and turbinate inflammation. Our findings suggest that the LN16-A strain is a promising live attenuated vaccine candidate that can prevent infection with subtype B aMPV.

Speech emotion recognition (SER) is a key technology in human-computer interaction (HCI) systems. Although the existing neural-based methods have achieved some satisfactory results in recognition accuracy, the failure of effective in-depth fusion of multi-scale features hinders the improvement of the accuracy of SER. In this paper, we address this issue from the two aspects of extracting exhaustive features and fusing features of multi-scale. In particular, we propose a recognition network based on Cross-Layer Intersectant Fusion, termed CLIF. It mainly consists of multi-scale feature extraction and cross-layer intersectant fusion. The former takes acoustic features as input and extracts feature maps with different receptive field ranges layer by layer through deepening convolution structures. Among these features, the lower level has more original information but also contains noise. The higher level has emotional semantics that is easier to classify but loses the perception of the details of the original acoustic features. Therefore, we use the cross-layer intersectant fusion module to achieve efficient utilization of low-level and high-level features. The experimental results demonstrate that the proposed CLIF is superior to the existing state-of-the-art speech emotion recognition algorithm. The overall recognition accuracies of CLIF can achieve 82.17% and 93.26% on IEMOCAP and CASIA datasets respectively.

In recent years, significant progress has been made in single-image super-resolution reconstruction (SR) for integer scale factors, but in many practical applications (such as image editing and visual enhancement) non-integer SR and asymmetric SR are required. However, the performance of such networks needs to be improved. In this paper, we propose an image arbitrary scale super-resolution reconstruction network based on the Cross-Scale Implicit Feature Sensing Module (CIFS). Our module could easily capture long-range spatial dependencies that can greatly improve network performance during image reconstruction. Networks with CIFS inserted will have better SR performance. In addition, CIFS has the characteristics that the input dimension is equal to the output dimension, which is easy to integrate into other networks.

Films restoration is a challenging yet promising task that involves using deep learning techniques to restore both structured and unstructured defects in old films. Significant progress has been made in this task by leveraging Vision Transformer and other existing technologies. However, there is still room for improvement in accurately localizing structured mask positions. Therefore, this paper proposes a novel module called the Mask-Aware (MA) module. By incorporating this module, our proposed model can achieve more precise mask positions and obtain higher-quality restoration results. Specifically, the MA module utilizes dynamic convolutions that employ attention mechanisms and deformable convolutions to enhance the ability of the model to perceive spatiotemporal information, thereby boosting the ability of the model to localize masks. Extensive experimental results demonstrate that our proposed model can achieve better handling of structural defects in old films.

Video frame interpolation (VFI) is a challenging yet promising task that involves synthesizing intermediate frames from two given frames. State-of-the-art approaches have made significant progress by directly synthesizing images using forward optical flow. However, these methods often encounter issues such as occlusion and pixel blurring when handling large motion scenes. To address these challenges, this paper proposes a novel approach based on the concept of local compensation. By adopting this approach, more refined optical flow estimation can be obtained, leading to higher-quality video frame interpolation results. Specifically, we introduce two modules, namely the Comprehensive Contextual Feature Extraction (CCFE) module and Motion-Guided Feature Fusion (MGFF) module, to enable local compensation of optical flow estimation. The CCFE module is designed to be embedded in each layer of the image pyramid structure. It aims to encourage the model to extract clean and sufficiently rich contextual information from the input images. On the other hand, the MGFF can guide the multi-source features fusion based on motion features, making the feature fusion of moving objects more precise, thus providing local compensation for optical flow estimation. Extensive experimental results demonstrate that incorporating our proposed modules into the baseline network significantly enhances the performance of video frame interpolation.

Frame-wise steganalysis is of significance for active steganography defense. By frame-wise detection, we can accurately find the embedding position of secret information and destroy the covert channel further. However, there is currently no research specifically aiming at frame-wise steganalysis of low-bit-rate compressed speech. Besides, most of the existing steganalysis methods are specifically designed for a specific category of steganography methods. They are difficult to apply to practical scenarios where the steganography algorithms are uncertain. In this paper, a general frame-wise steganalysis method for low-bit-rate compressed speech is proposed. To extract rich feature from a speech frame, we propose a dual-domain representation, which conducts feature extraction both in the compressed domain and the decoded time domain. In addition, we propose an efficient steganalysis network named Stegaformer to leach the intra-frame correlation from the obtained representation to enable steganalysis. In Stegaformer, an adaptive local correlation enhancement module is introduced to effectively models the local characteristics, which compensates for the drawback of traditional Transformer-based models. Experimental results show that our method performs better than the existing steganalysis methods in detecting multiple steganography methods for a speech frame.

旨在确定我国部分养鸡场的鸡出现甩头、精神萎靡和肿头综合征是否由禽偏肺病毒(aMPV)引起,本研究从山东、福建、黑龙江等地的发病蛋鸡和肉鸡场采集鼻甲骨、气管和肺等样品,首先利用aMPV特异性的RT-PCR方法对临床样品进行初步检测,将RT-PCR检测阳性样品接种Vero细胞进行病毒分离,然后利用G/F基因序列分析及间接免疫荧光试验(IFA)等鉴定病毒的亚型,最后将分离株感染SPF鸡进行致病性分析.结果显示:在采集的220份样品中,RT-PCR检测结果显示,有3份鼻甲骨样品在228 bp左右出现特异性条带,将阳性病料接种Vero细胞盲传5代后,细胞出现变圆、聚集和融合等aMPV特征性细胞病变(CPE),表明成功分离到3株aMPV,将其分别命名为SD2001、SD2002和HLJ2101.G和F基因同源性分析显示,来自蛋鸡的SD2001、SD2002和来自肉鸡的HLJ2101分离株的G和F基因与其他B亚型aMPV毒株的核苷酸和氨基酸序列相似性均较高,核苷酸的相似性分别为93.4％～98.6％和95.6％～100.0％,氨基酸的相似性分别为88.7％～97.8％和97.6％～100.0％;而与A、C和D亚型的G和F基因同源性较低,核苷酸的相似性分别为27.1％～61.8％和66.8％～74.8％,氨基酸的相似性分别为16.1％～36.7％和72.5％～86.5％,这些结果表明,SD2001、SD2002和HLJ2101分离株属于B亚型aMPV.进一步利用B亚型aMPV特异性的阳性血清进行IFA检测,接种SD2001、SD2002和HLJ2101的Vero细胞均可以观察到特异性的绿色荧光信号,进一步证实3个分离株属于B亚型aMPV.选择SD2001感染3周龄SPF鸡进行了致病性研究,结果发现SPF鸡感染后3～6 d出现精神萎靡、甩头和流鼻涕等症状,鼻甲骨、气管和肺也出现病理性损伤,其发病率为90％(18/20).3株B亚型aMPV的分离不仅有助于明确我国部分养鸡场出现肿头综合征的发病原因,同时也证实B亚型aMPV流行毒株对鸡有明显的致病性,这些结果为我国家禽疫病的诊断和有效防控提供了重要理论依据.

Primates are facing a serious survival crisis. Tracking the range of animal activities and population changes is of great significance for efficient animal protection. Primates are highly alert and inaccessible to humans so that it is difficult to track animals through direct observation, DNA fingerprinting, or marking methods. Primate recognition based on animal calls has the advantages of wide monitoring range, low equipment cost, and good concealment. In this work, we propose an effective macaque speech feature extraction structure, and innovatively propose a feature fusion mechanism to effectively obtain the feature representation of each call. Furthermore, we construct a public open source macaque voiceprint verification dataset. The experimental results show that the proposed method is superior to the existing state-of-the-art human voiceprint verification algorithms with different call durations. The equal error rate (EER) of our macaque voiceprint verification algorithm reaches 6.19%.

In this paper, a frame-level steganalysis of Quantization Index Modulation (QIM) steganography in compressed speech streams is proposed for the first time. The proposed method builds a neural network classification framework based on multi-dimensional perspective of codeword correlations, which is inspired by cognitive biology. Four dimensions are employed: global-to-local, local-to-global, forward and backward. First, the codeword embedding method is utilized to map each codeword into a compact representation. Next, Bi-LSTM is used to consider the steganographic features in time sequence and reverse time sequence. Subsequently, a dual-thread attention mechanism is designed to extract local and global features at the same time. Finally, a channel attention mechanism is employed to increase the weight that contributes the most to the current task and the convolution and fully connected layers are used to generate the frame-level steganographic label. Experimental results show that the proposed method is effective and practical in frame-level detection tasks.

为研究B亚型禽偏肺病毒(aMPV/B)对蛋鸡的致病性,将aMPV/B LN16株经Vero细胞传代培养,通过点眼滴鼻的方式感染9周龄蛋鸡.感染后第2天,感染组蛋鸡开始出现流鼻涕,并伴有鼻痂等症状,发病率为92.3％.荧光定量PCR检测结果表明,感染组蛋鸡在感染后1～5 d排毒.感染后第4天,感染组蛋鸡的哈氏腺、气管、喉头及鼻甲骨内均有病毒分布.感染后7～21d,感染组蛋鸡的aMPV抗体水平逐渐升高.组织病理观察结果表明,感染组蛋鸡的鼻甲骨、肺脏、气管均表现出不同程度的病理损伤,主要表现为炎性细胞浸润.研究表明,aMPV/B可以感染蛋鸡并引起发病.

Steganographic algorithms in low-bit-rate compressed speech bring convenience to realize covert communication, meanwhile result in safety issues. The existing steganalysis methods are normally designed for one specific category of steganographic methods, thus lacking generalization capability. In this paper, we propose a general steganalysis method based on code element (CE) embedding, Bi-LSTM and CNN with attention mechanisms. Firstly, CEs in each frame are converted to a multi-hot vector. And each multi-hot vector will be mapped into a fixed-length embedding vector to get a more compact representation by utilizing dictionaries. Then, Bi-LSTM and CNN are applied to extract the contextual information and the local characteristics respectively of these embedding vectors. In addition, the attention mechanisms are introduced in different layers of the network to assign different weights to the output feature within each layer. Finally, the prediction results can be generated by the fully connected layer. Experimental results show that our method performs better than the existing steganalysis methods for detecting multiple steganography methods in the low-bit-rate compressed speech streams.

Analysis-by-synthesis linear predictive coding (AbS-LPC) is widely used in a variety of low-bit-rate speech codecs. Most of the current steganalysis methods for AbS-LPC low-bit-rate compressed speech steganography are specifically designed for a specific coding standard or category of steganography methods, and thus lack generalization capability. In this paper, a general steganalysis method for detecting steganographies in low-bit-rate compressed speech under different standards is proposed. First, the code-element matrices corresponding to different coding standards are concatenated to obtain a synthetic code-element matrix, which will be mapped into an intermediate feature representation by utilizing the pre-trained dictionaries. Then, bidirectional long short-term memory is employed to capture long-term contextual correlations. Finally, a code-element affinity attention mechanism is used to capture the global inter-frame context, and a full connection structure is used to generate the prediction result. Experimental results show that the proposed method is effective and better than the comparison methods for detecting steganographies in cross-standard low-bit-rate compressed speech.

In this paper, we present a universal steganalysis method for both intra prediction mode and motion vector-based steganography based on deep learning. Since the embedding process is eventually reflected in the modification of pixel values in decoded frames, we design a Noise Residual Convolutional Neural Network (NR-CNN) from the perspective of the spatial domain, which is the first CNN-based approach for this subject. In NR-CNN, feature extraction and classification modules are integrated into a unified and trainable network framework. It automatically learns features and implements classification in a data-driven manner, which effectively solves the existing problems. Experimental results show that NR-CNN has better performance of steganalysis than the related method.

天气状况对室外视频设备的成像效果有很大影响。为实现成像设备在恶劣天气下的自适应调整,从而提升智能监控系统的效果,同时针对传统的天气图像判别方法分类效果差且对相近天气现象不易分类的不足,以及深度学习方法识别天气准确率不高的问题,提出了一个将传统方法与深度学习方法相结合的特征融合模型。融合模型采用4种人工设计算法提取传统特征,采用AlexNet提取深层特征,利用融合后的特征向量进行图像天气状况的判别。融合模型在多背景数据集上的准确率达到93. 90%,优于对比的3种常用方法,并且在平均精准率（AP）和平均召回率（AR）指标上也表现良好;在单背景数据集上的准确率达到96. 97%,AP和AR均优于其他模型,且能很好识别特征相近的天气图像。实验结果表明提出的特征融合模型可以结合传统方法和深度学习方法的优势,提升现有天气图像分类方法的准确度,同时提高在特征相近的天气现象下的识别率。

Recent progress in vision-based fire detection is driven by convolutional neural networks. However, the existing methods fail to achieve a good tradeoff among accuracy, model size, and speed. In this paper, we propose an accurate fire detection method that achieves a better balance in the abovementioned aspects. Specifically, a multiscale feature extraction mechanism is employed to capture richer spatial details, which can enhance the discriminative ability of fire-like objects. Then, the implicit deep supervision mechanism is utilized to enhance the interaction among information flows through dense skip connections. Finally, a channel attention mechanism is employed to selectively emphasize the contribution between different feature maps. Experimental results demonstrate that our method achieves 95.3% accuracy, which outperforms the suboptimal method by 2.5%. Moreover, the speed and model size of our method are 3.76% faster on the GPU and 63.64% smaller than the suboptimal method, respectively.

In recent years, with the widespread application of encryption technology, criminals can hide malicious data without being discovered by security regulatory authorities, which has brought serious challenges to computer forensic investigation. Therefore, it is urgent to study the technology of detection and forensics of encrypted data. This paper proposes a method for encryption detection based on a deep convolutional neural network. The method first converts the raw data into two-dimensional matrixes as the input of the convolutional neural network. Then, the multiscale feature extraction mechanism with multiple activation functions is utilized to provide representative features as the input of subsequent layers. Next, the residual learning operation can further enhance the discrimination of features. By this mean, a network which can automatically extract and learn global contextual information of encrypted data is constructed. The experiment results show that the proposed method achieves high accuracy in the detection of storage file and network transmission data compare to the competitive methods and the detection accuracy on different types of mixed data is higher than 99%. Moreover, the proposed method can accurately detect data encrypted with different algorithms. The average detection rate of DES-encrypted data is higher than that of competitors by more than 5%.

Analysis-by-synthesis linear predictive coding (AbS-LPC) is widely used in a variety of low-bit-rate speech codecs. The existing steganalysis methods for AbS-LPC low-bit-rate compressed speech steganography are specifically designed for one certain category of steganography methods, thus lacking generalization capability. In this paper, a common method for detecting multiple steganographies in low-bit-rate compressed speech based on a code element Bayesian network is proposed. In an AbS-LPC low-bit-rate compressed speech stream, spatiotemporal correlations exist between the code elements, and steganography will eventually change the values of these code elements. Thus, the method presented in this paper is developed from the code element perspective. It consists of constructing a code element Bayesian network based on the strong correlations between code elements, learning the network parameters by utilizing a Dirichlet distribution as the prior distribution, and finally implementing steganalysis based on Bayesian inference. Experimental results demonstrate that the proposed method performs better than the existing steganalysis methods for detecting multiple steganographies in the AbS-LPC low-bit-rate compressed speech.

In this paper, we focus on quantization-index-modulation (QIM) steganography in low-bit-rate speech codec and contribute to improve its steganalysis resistance. A novel QIM steganography is proposed based on the replacement of quantization index set in linear predictive coding (LPC). In this method, each quantization index set is seen as a point in quantization index space. Steganography is conducted in such space. Comparing with other methods, our algorithm significantly improves the embedding efficiency. One quantization index needs to be changed at most when three binary bits are hidden. The number of alterations introduced by the proposed approach is much lower than that of the current methods with the same embedding rate. Due to the fewer cover changes, the proposed steganography is less detectable. Moreover, a division strategy based on the genetic algorithm is proposed to reduce the additional distortion introduced by replacements. In our experiment, ITU-T G.723.1 is selected as the codec, and the experimental results show that the proposed approach outperforms the state-of-the-art LPC-based approach in low-bit-rate speech codec with respect to both steganographic capacity and steganalysis resistance.