Reliable recognition of composite radar jamming signals under severe noise conditions remains a critical challenge in modern electronic warfare. To address the feature drowning issue in low jamming-to-noise ratio (JNR) environments, we propose a deep learning-based dual-branch network that integrates a shifted-window Transformer branch and a convolutional branch (Swin-Conv) through a concatenation-based selective kernel (CSK) fusion module, termed SC-CSKNet. The proposed network jointly explores the local transient structures and global modulation dependencies of the jamming signal. Moreover, a residual bi-directional cross-attention module is designed for feature alignment, while a concatenation-based selective kernel fusion (CSK-fusion) module is used as an adaptive gate to dynamically suppress the noise-corrupted channels. Extensive experiments demonstrate that SC-CSKNet achieves a superior average recognition accuracy of 72.65% even in the low-JNR region and outperforms the best baseline ConvNeXt Tiny by 2.32%, while maintaining excellent computational efficiency.

To ensure stringent security strategies for image information involving personal privacy during conveyance and storage, we propose an innovative multiface privacy protection scheme based on chaos theory. Compared to single-face encryption algorithms, the proposed scheme has broader potential applications in fields, such as smart cities and smart transportation. Specifically, a new spatiotemporal chaotic system named the sine-cosine coupled mapping lattice system (SCCML) is designed. It features a larger parameter domain, complexity, and profound unpredictability, yet maintains a simpler construction aimed at providing potential benefits and implementations in the field of information security. In the proposed multiface privacy protection scheme, multiple faces within an image are rapidly and accurately identified and then encrypted using the proposed SCCML-based digital separation loop encryption algorithm. The encryption algorithm exhibits a synchronous scrambling diffusion mechanism. Additionally, the introduction of mixed multibase cascade diffusion offers multiple layers of security for facial data, prevents diffusion singularity, and enhances diversity, making it significantly more challenging to crack. Experimental verification on a real multiface image dataset shows that the algorithm is superior, practical, safe, and efficient.

Deep learning based techniques have been popularly adopted in acoustic echo cancellation (AEC). Utilization of speaker representation has extended the frontier of AEC, thus attracting many researchers’ interest in personalized acoustic echo cancellation (PAEC). Meanwhile, task-decoupling strategies are widely adopted in speech enhancement. To further explore the task-decoupling approach, we propose to use a two-stage task-decoupling post-filter (TDPF) in PAEC. Furthermore, a multi-scale local-global speaker representation is applied to improve speaker extraction in PAEC. Experimental results indicate that the task-decoupling model can yield better performance than a single joint network. The optimal approach is to decouple the echo cancellation from noise and interference speech suppression. Based on the task-decoupling sequence, optimal training strategies for the two-stage model are explored afterwards.

This paper describes our NPU-Elevoc personalized speech enhancement system (NAPSE) for the 5th Deep Noise Suppression Challenge[1] at ICASSP 2023. Based on the superior two-stage model TEA-PSE 2.0 [2], our system particularly explores better strategy for speaker embedding fusion, optimizes the model training pipeline, and leverages adversarial training and multi-scale loss. According to the results12, our system is tied for the 1st place in the headset track (track 1) and ranked 2nd in the speakerphone track (track 2).

This paper describes Tencent Ethereal Audio Lab – Northwestern Polytechnical University personalized speech enhancement (TEA-PSE) system submitted to track 2 of the ICASSP 2022 Deep Noise Suppression (DNS) challenge. Our system specifically combines the dual-stage network which is a superior real-time speech enhancement framework with the ECAPA-TDNN speaker embedding network which achieves state-of-the-art performance in speaker verification. The dual-stage network aims to decouple the primal speech enhancement problem into multiple easier sub-problems. Specifically, in stage 1, only the magnitude of the target speech is estimated, which is incorporated with the noisy phase to obtain a coarse complex spectrum estimation. To facilitate the formal estimation, in stage 2, an auxiliary network serves as a post-processing module, where residual noise and interfering speech are further suppressed and the phase information is effectively modified. With the asymmetric loss function to penalize over-suppression, more target speech is preserved, which is helpful for both speech recognition performance and subjective sense of hearing. Our system reaches 3.97 in overall audio quality (OVRL) MOS and 0.69 in word accuracy (WAcc) on the blind test set of the challenge, which outperforms the DNS baseline by 0.57 OVRL and ranks 1st in track 2.

Chaotic image encryption is widely used in the field of information security. This paper proposes a novel chaotic image encryption method with spiral-transform-based fractal sorting matrix (STFSM). First of all, the theory of STFSM with good scrambling effect is introduced, which has good irregularity and iterative. Then, the iterative algorithm and calculation example of STFSM are introduced. STFSM can be used as the map of spatial location transformations to implement the design of image encryption. Based on the complete STFSM theory and iterative algorithm, a chaotic image cryptosystem based on STFSM is proposed to achieve a good image encryption process. To test the security of the proposed algorithm, security tests and analyses such as entropy analysis, correlation analysis, resistance to differential attacks analysis, and robustness analysis are used for the proposed algorithm. The experimental analysis illustrates that the algorithm has a better encryption effect, whether using conventional tests or the attacks simulations described in this paper and can also effectively resist the attacks.

Current out-of-distribution (OOD) detection benchmarks are commonly built by defining one dataset as in-distribution (ID) and all others as OOD. However, these benchmarks unfortunately introduce some unwanted and impractical goals, e.g., to perfectly distinguish CIFAR dogs from ImageNet dogs, even though they have the same semantics and negligible covariate shifts. These unrealistic goals will result in an extremely narrow range of model capabilities, greatly limiting their use in real applications. To overcome these drawbacks, we re-design the benchmarks and propose the semantically coherent out-of-distribution detection (SC-OOD). On the SC-OOD benchmarks, existing methods suffer from large performance degradation, suggesting that they are extremely sensitive to low-level discrepancy between data sources while ignoring their inherent semantics. To develop an effective SC-OOD detection approach, we leverage an external unlabeled set and design a concise framework featured by unsupervised dual grouping (UDG) for the joint modeling of ID and OOD data. The proposed UDG can not only enrich the semantic knowledge of the model by exploiting unlabeled data in an unsupervised manner, but also distinguish ID/OOD samples to enhance ID classification and OOD detection tasks simultaneously. Extensive experiments demonstrate that our approach achieves the state-of-the-art performance on SC-OOD benchmarks. Code and benchmarks are provided on our project page: https://jingkang50.github.io/projects/scood.

Webly supervised learning becomes attractive recently for its efficiency in data expansion without expensive human labeling. However, adopting search queries or hashtags as web labels of images for training brings massive noise that degrades the performance of DNNs. Especially, due to the semantic confusion of query words, the images retrieved by one query may contain tremendous images belonging to other concepts. For example, searching 'tiger cat' on Flickr will return a dominating number of tiger images rather than the cat images. These realistic noisy samples usually have clear visual semantic clusters in the visual space that mislead DNNs from learning accurate semantic labels. To correct real-world noisy labels, expensive human annotations seem indispensable. Fortunately, we find that metadata can provide extra knowledge to discover clean web labels in a labor-free fashion, making it feasible to automatically provide correct semantic guidance among the massive label-noisy web data. In this paper, we propose an automatic label corrector VSGraph-LC based on the visual-semantic graph. VSGraph-LC starts from anchor selection referring to the semantic similarity between metadata and correct label concepts, and then propagates correct labels from anchors on a visual graph using graph neural network (GNN). Experiments on realistic webly supervised learning datasets Webvision-1000 and NUS-81-Web show the effectiveness and robustness of VSGraph-LC. Moreover, VSGraph-LC reveals its advantage on the open-set validation set.

Resembling the rapid learning capability of human, low-shot learning empowers vision systems to understand new concepts by training with few samples. Leading approaches derived from meta-learning on images with a single visual object. Obfuscated by a complex background and multiple objects in one image, they are hard to promote the research of low-shot object detection/segmentation. In this work, we present aflexible and general methodology to achieve these tasks. Our work extends Faster /Mask R-CNN by proposing meta-learning over RoI (Region-of-Interest) features instead of a full image feature. This simple spirit disentangles multi-object information merged with the background, without bells and whistles, enabling Faster /Mask R-CNN turn into a meta-learner to achieve the tasks. Specifically, we introduce a Predictor-head Remodeling Network (PRN) that shares its main backbone with Faster /Mask R-CNN. PRN receives images containing low-shot objects with their bounding boxes or masks to infer their class attentive vectors. The vectors take channel-wise soft-attention on RoI features, remodeling those R-CNN predictor heads to detect or segment the objects consistent with the classes these vectors represent. In our experiments, Meta R-CNN yields the new state of the art in low-shot object detection and improves low-shot object segmentation byMaskR-CNN.Code: https://yanxp.github.io/metarcnn.html.

A broad range of cross-$m$-domain generation researches boil down to matching a joint distribution by deep generative models (DGMs). Hitherto algorithms excel in pairwise domains while as $m$ increases, remain struggling to scale themselves to fit a joint distribution. In this paper, we propose a domain-scalable DGM, i.e., MMI-ALI for $m$-domain joint distribution matching. As an $m$-domain ensemble model of ALIs (Dumoulin et al., 2016), MMI-ALI is adversarially trained with maximizing Multivariate Mutual Information (MMI) w.r.t. joint variables of each pair of domains and their shared feature. The negative MMIs are upper bounded by a series of feasible losses provably leading to matching $m$-domain joint distributions. MMI-ALI linearly scales as $m$ increases and thus, strikes a right balance between efficacy and scalability. We evaluate MMI-ALI in diverse challenging $m$-domain scenarios and verify its superiority.

Though quite challenging, leveraging large-scale unlabeled or partially labeled data in learning systems (e.g., model/classifier training) has attracted increasing attentions due to its fundamental importance. To address this problem, many active learning (AL) methods have been proposed that employ up-to-date detectors to retrieve representative minority samples according to predefined confidence or uncertainty thresholds. However, these AL methods cause the detectors to ignore the remaining majority samples (i.e., those with low uncertainty or high prediction confidence). In this paper, by developing a principled active sample mining (ASM) framework, we demonstrate that cost-effective mining samples from these unlabeled majority data are a key to train more powerful object detectors while minimizing user effort. Specifically, our ASM framework involves a switchable sample selection mechanism for determining whether an unlabeled sample should be manually annotated via AL or automatically pseudolabeled via a novel self-learning process. The proposed process can be compatible with mini-batch-based training (i.e., using a batch of unlabeled or partially labeled data as a one-time input) for object detection. In this process, the detector, such as a deep neural network, is first applied to the unlabeled samples (i.e., object proposals) to estimate their labels and output the corresponding prediction confidences. Then, our ASM framework is used to select a number of samples and assign pseudolabels to them. These labels are specific to each learning batch based on the confidence levels and additional constraints introduced by the AL process and will be discarded afterward. Then, these temporarily labeled samples are employed for network fine-tuning. In addition, a few samples with low-confidence predictions are selected and annotated via AL. Notably, our method is suitable for object categories that are not seen in the unlabeled data during the learning process. Extensive experiments on two public benchmarks (i.e., the PASCAL VOC 2007/2012 data sets) clearly demonstrate that our ASM framework can achieve performance comparable to that of the alternative methods but with significantly fewer annotations.

Though quite challenging, leveraging large-scale unlabeled or partially labeled images in a cost-effective way has increasingly attracted interests for its great importance to computer vision. To tackle this problem, many Active Learning (AL) methods have been developed. However, these methods mainly define their sample selection criteria within a single image context, leading to the suboptimal robustness and impractical solution for large-scale object detection. In this paper, aiming to remedy the drawbacks of existing AL methods, we present a principled Self-supervised Sample Mining (SSM) process accounting for the real challenges in object detection. Specifically, our SSM process concentrates on automatically discovering and pseudo-labeling reliable region proposals for enhancing the object detector via the introduced cross image validation, i.e., pasting these proposals into different labeled images to comprehensively measure their values under different image contexts. By resorting to the SSM process, we propose a new AL framework for gradually incorporating unlabeled or partially labeled data into the model learning while minimizing the annotating effort of users. Extensive experiments on two public benchmarks clearly demonstrate our proposed framework can achieve the comparable performance to the state-of-the-art methods with significantly fewer annotations.

In this study, non‐orthogonal multiple access (NOMA)‐based multiple‐antenna and multiple‐relay networks over Nakagami‐ m fading channels with imperfect channel state information (CSI) and successive interference cancellation (SIC) error are investigated. In order to measure the performance of the system, the exact outage probability and the ergodic sum rate of the system are explored. By analysing the impact of the partial relay selection scheme, it is shown that the performance gain can be achieved significantly by increasing the number of relays from one to two, but when the number of relays is equal to or greater than three, the increment of the performance gain becomes smaller and smaller, and finally saturated as the number of relays increases. Furthermore, due to the complexity of the derivation process for the ergodic sum rate, an upper bound of the ergodic sum rate is taken into account. Finally, by carrying out a series of experimental simulations, the authors prove the correctness of the theoretical derivations above, and from the comparison of NOMA and orthogonal multiple access (OMA), it is found that NOMA can provide better spectral efficiency and user fairness than OMA.

To satisfy the requirement of practical applications, many certificateless encryption schemes (CLE) without pairing have been proposed. Recently, Lai et al . proposed a CLE scheme without pairing and demonstrated that their scheme is provably secure in the random oracle model. The analysis shows that their scheme has better performance than the related schemes. However, Lai et al . ’s scheme is not a standard CLE scheme since the user's public key is used when generating his partial private key. In this study, the authors propose a new CLE scheme. Compared with Lai et al .’s scheme, the authors' scheme is a standard CLE scheme at the cost of increasing the computational cost slightly. Besides, their scheme has better performance than the related schemes except Lai et al .’s scheme. They also show their scheme is provably secure in the random oracle model.