{"title":"Domain-Assisted Few-Shot Linguistic Steganalysis in Imbalanced Class Scenarios","authors":"Qingying Niu;Zhen Yang;Yufei Luo;Jiangrui Zhao;Yuwen Jiang","doi":"10.1109/LSP.2025.3553427","DOIUrl":"https://doi.org/10.1109/LSP.2025.3553427","url":null,"abstract":"Linguistic steganalysis aims to distinguish stego text from cover text. However, most existing methods heavily rely on a large number of stego text samples for training. In real-world scenarios, the cover text is far more abundant than the stego text, making it extremely difficult to obtain sufficient stego text for training. Furthermore, the scarcity of stego text also increases the difficulty of detection, posing greater challenges for steganalysis. In contrast, cover text is relatively easier to obtain in real-world scenarios, but current methods fail to fully utilize this resource. In this paper, we propose a Domain-Assisted Few-shot linguistic steganalysis method called DAF-Stega. To make full use of the cover text, we incorporate cover texts from multiple domains to assist in training. To address the scarcity of stego texts, we perform few-shot steganalysis based on a small amount of stego text and employ dynamic decision-making to generate pseudo-labels for self-training, enhancing model performance. Experimental results show that in few-shot learning scenarios, DAF-Stega effectively addresses the steganalysis problem under uncertain stego text proportions and outperforms existing methods.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"32 ","pages":"1391-1395"},"PeriodicalIF":3.2,"publicationDate":"2025-03-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748911","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Dual-Branch Network for No-Reference Super-Resolution Image Quality Assessment","authors":"Tong Tang;Fan Yang;Xinyu Lin;Weisheng Li","doi":"10.1109/LSP.2025.3553432","DOIUrl":"https://doi.org/10.1109/LSP.2025.3553432","url":null,"abstract":"No-reference super-resolution image quality assessment (SR-IQA) has become an critical technique for optimizing SR algorithms, the key challenge is how to comprehensively learn visual related features of SR image. Existing methods ignore the context information and feature correlation. To tackle this problem, this letter proposes a dual-branch network for no-reference super-resolution image quality assessment (DBSRNet). First, dual-branch feature extraction module is designed, where residual network and receptive field block net are combined to learn multi-scale local features, stacked vision transformer blocks are utilized to learn global features. Then, correlations between dual-branch features are learned and fused based on self-attention mechanism structure, final predicted score is obtained by adaptive feature pooling strategy. Finally, experimental results show that DBSRNet significantly outperforms State-of-the-Art methods in terms of prediction accuracy on all SR-IQA datasets.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"32 ","pages":"1366-1370"},"PeriodicalIF":3.2,"publicationDate":"2025-03-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143726419","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Modified Price Nonlinear Frequency Modulated Waveform for Improved Performance","authors":"Geon U Kim;Jeong Phill Kim","doi":"10.1109/LSP.2025.3553437","DOIUrl":"https://doi.org/10.1109/LSP.2025.3553437","url":null,"abstract":"A modified Price formula nonlinear frequency curves for pulse compression with ultra-low sidelobe level (SLL) and efficient use of the frequency spectrum is proposed. Although the frequency curve of the original Price formula is known to give a relatively low remaining SLL, this level is not suitable for applications such as meteorological radar, which needs to detect tiny targets in a dense and widespread cluttered environment. This paper introduces new parameters for obtaining an ultra-low SLL. After a parameter study, hybrid optimization based on the genetic algorithm and the Nelder-Mead search was used for finding the optimal solution. In this method, even without amplitude windowing, an SLL lower than <inline-formula><tex-math>$-100$</tex-math></inline-formula> dB could be achieved with time-bandwidth product of 200. In addition, another design with a one-sided occupied spectrum (<inline-formula><tex-math>$ f_{-90text{dB}}$</tex-math></inline-formula>) of only 32.85 MHz and a Doppler tolerance of <inline-formula><tex-math>$-0.24$</tex-math></inline-formula> dB was made.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"32 ","pages":"1406-1410"},"PeriodicalIF":3.2,"publicationDate":"2025-03-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748787","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"SequenceOut: Boosting CNNs by Freezing Layers","authors":"Shitala Prasad;Rakesh Paul;Mayur Kamat","doi":"10.1109/LSP.2025.3553430","DOIUrl":"https://doi.org/10.1109/LSP.2025.3553430","url":null,"abstract":"Convolutional neural networks (CNNs) are a powerful tool for various computer vision tasks, demonstrating exceptional performance in image classification, object detection, and segmentation. However, traditional training methods often require meticulous hyperparameter tuning, architectural adjustments, or the introduction of additional data through techniques such as data augmentation to achieve optimal accuracy. This letter introduces an innovative training strategy that leverages layer freezing to enhance the training process while keeping the model's architecture and hyperparameters unchanged. By selectively and progressively freezing certain hidden layers in the CNN, we prevent the model from reaching a saturation point. This approach effectively reduces the backpropagation parameter space, facilitating more focused and efficient learning in the remaining layers.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"32 ","pages":"1401-1405"},"PeriodicalIF":3.2,"publicationDate":"2025-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748786","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"ViLNM: Visual-Language Noise Modeling for Text-to-Image Person Retrieval","authors":"Guolin Xu;Yong Feng;Yanying Chen;Guofan Duan;Mingliang Zhou","doi":"10.1109/LSP.2025.3553424","DOIUrl":"https://doi.org/10.1109/LSP.2025.3553424","url":null,"abstract":"Text-to-image person retrieval (TPR) focuses on finding a specific person based on the textual description, and most methods implicitly assume the training image-text pairs are correctly aligned. In practice, the image-text pairs exist under-correlated or false-correlated due to the low quality of the images and annotation errors. Meanwhile, remarkable similarities between different person identities may lead to a mismatch between text and image. To tackle the two issues, we present a Visual-Language Noise Modeling (ViLNM) method that successfully captures robust cross-modal associations even with noise. Specifically, we design a Noise Token Aware (NTA) module that eliminates the words in the textual description that do not match the image, utilizing the matched words to establish a more reliable association. Besides, to enhance the recognition ability of the model for different person identities, we propose a Joint Inter and Intra-Modal Contrastive Loss (JII) and Local Aggregation (LA) module to increase the feature differences between different person identities. We conduct comprehensive experiments on three public benchmarks, and ViLNM performs best.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"32 ","pages":"1386-1390"},"PeriodicalIF":3.2,"publicationDate":"2025-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748913","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Low Complexity MRC Detection of IRS-Aided Single User MIMO-OTFS","authors":"Sapta Girish Neelam","doi":"10.1109/LSP.2025.3553428","DOIUrl":"https://doi.org/10.1109/LSP.2025.3553428","url":null,"abstract":"This paper presents a novel detection strategy for Intelligent Reflecting Surface (IRS)-aided single user Multiple-Input Multiple-Output (MIMO) systems utilizing Orthogonal Time Frequency Space (OTFS) modulation, tailored to operate effectively under hardware constraints such as Carrier Frequency Offset (CFO). The proposed method employs Maximum Ratio Combining (MRC) to enhance signal quality by mitigating multipath fading and inter-antenna interference. A notable feature of this detection strategy is its low computational complexity, which makes it highly practical for real-time applications in dynamic wireless environments. Designed for low computational complexity, this detection scheme significantly improves the performance of IRS-aided MIMO-OTFS systems. Simulation results demonstrate the superior capabilities of the proposed approach, as IRS-aided single-user MIMO-OTFS systems with MRC detection consistently outperform traditional detection methods. These findings highlight the transformative potential of integrating IRS and OTFS to advance the next generation of wireless communication systems.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"32 ","pages":"1396-1400"},"PeriodicalIF":3.2,"publicationDate":"2025-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748914","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Blind Deconvolution of Graph Signals: Robustness to Graph Perturbations","authors":"Chang Ye;Gonzalo Mateos","doi":"10.1109/LSP.2025.3553064","DOIUrl":"https://doi.org/10.1109/LSP.2025.3553064","url":null,"abstract":"We study blind deconvolution of signals defined on the nodes of an undirected graph. Although observations are bilinear functions of both unknowns, namely the forward convolutional filter coefficients and the graph signal input, a filter invertibility requirement along with input sparsity allow for an efficient linear programming reformulation. Unlike prior art that relied on perfect knowledge of the graph eigenbasis, here we derive stable recovery conditions in the presence of small graph perturbations. We also contribute a provably convergent robust algorithm, which alternates between blind deconvolution of graph signals and eigenbasis denoising in the Stiefel manifold. Reproducible numerical tests showcase the algorithm's robustness under several graph eigenbasis perturbation models.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"32 ","pages":"1381-1385"},"PeriodicalIF":3.2,"publicationDate":"2025-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748912","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Factorization-Based Information Reconstruction for Enhancing Missing Modality Robustness","authors":"Chao Wang;Miaolei Zhou;Xiaofei Yu;Junchao Weng;Yong Zhang","doi":"10.1109/LSP.2025.3552435","DOIUrl":"https://doi.org/10.1109/LSP.2025.3552435","url":null,"abstract":"In recent years, Multimodal Sentiment Analysis (MSA) has emerged as a prominent research area, utilizing multiple signals to better understand human sentiment. Previous studies in MSA have primarily concentrated on performing interaction and fusion with complete signals. However, they have overlooked the issue of missing signals, which commonly arise in real-world scenarios due to factors such as occlusion, privacy concerns, and device malfunctions, leading to reduced generalizability. To this end, we propose a Factorization-based Information Reconstruction Framework (FIRF) to mitigate the modality missing problem in the MSA task. Specifically, we propose a fine-grained complementary factorization module that factorizes modality into synergistic, modality-heterogeneous, and noisy representations and design elaborate constraint paradigms for representation learning. Furthermore, we design a distribution calibration self-distillation module that fully recovers the missing semantics by utilizing bidirectional knowledge transfer. Comprehensive experiments on two datasets indicate that FIRF has a significant performance advantage over previous methods with uncertain missing modalities.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"32 ","pages":"1376-1380"},"PeriodicalIF":3.2,"publicationDate":"2025-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143748819","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"DNGG: Medical Image Lossless Encryption via Deep Network Guided Generative","authors":"Lin Fan;Meng Li;Zhenting Hu;Yuan Hong;Dexing Kong","doi":"10.1109/LSP.2025.3552528","DOIUrl":"https://doi.org/10.1109/LSP.2025.3552528","url":null,"abstract":"Ensuring the security and integrity of medical images is crucial for telemedicine. Recently, deep learning-based image encryption techniques have significantly improved data transmission security. However, the unpredictability of complex models may lead to damage during image reconstruction, thereby negatively impacting medical diagnostics. To address this issue, we propose a lossless encryption algorithm for medical images, which is based on a guided image generative neural network. Initially, we designed a guided image generation network. Subsequently, we train a generator using random keys to produce a key map. This key map then guides the encryption of the secret image through a bitwise XOR (bit-XOR) algorithm, effectively merging the secret image with the key map. During the decryption process, the original image can be restored losslessly by using a key map generated from a random key. The experimental results show that the encryption algorithm greatly ensures the security of data and shows strong anti-attack ability.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"32 ","pages":"1331-1335"},"PeriodicalIF":3.2,"publicationDate":"2025-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143716487","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Optimizing the Order of Modes in Tensor Train Decomposition","authors":"Petr Tichavský;Ondřej Straka","doi":"10.1109/LSP.2025.3552005","DOIUrl":"https://doi.org/10.1109/LSP.2025.3552005","url":null,"abstract":"The tensor train (TT) is a popular way of representing high-dimensional hyper-rectangular data structures called tensors. It is widely used, for example, in quantum chemistry under the name “matrix product state”. The complexity of the TT model mainly depends on the bond dimensions that connect TT cores, constituting the model. Unlike canonical polyadic decomposition, the TT model complexity may depend on the order of the modes/indices in the data structures or the order of the core tensors in the TT, in general. This letter aims to provide methods for optimizing the order of the modes to reduce the bond dimensions. Since the number of possible orderings of the cores is exponentially high, we propose a greedy algorithm that provides a suboptimal solution. We consider three problem setups, i.e., specifications of the tensor: tensor given by a list of all its elements, tensor described by a TT model with some default order of the modes, and tensor obtained by sampling a multivariate function.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"32 ","pages":"1361-1365"},"PeriodicalIF":3.2,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143725106","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}