Shugong Xu

Xiaoxiang Han, Xinyu Li, Jiang Shang et al.

Segmenting anatomical structures and lesions from ultrasound images contributes to disease assessment. Weakly supervised learning (WSL) based on sparse annotation has achieved encouraging performance and demonstrated the potential to reduce annotation costs. This study attempts to introduce scribble-based WSL into ultrasound image segmentation tasks. However, ultrasound images often suffer from poor contrast and unclear edges, coupled with insufficient supervison signals for edges, posing challenges to edge prediction. Uncertainty modeling has been proven to facilitate models in dealing with these issues. Nevertheless, existing uncertainty estimation paradigms are not robust enough and often filter out predictions near decision boundaries, resulting in unstable edge predictions. Therefore, we propose leveraging predictions near decision boundaries effectively. Specifically, we introduce Dempster-Shafer Theory (DST) of evidence to design an Evidence-Guided Consistency strategy. This strategy utilizes high-evidence predictions, which are more likely to occur near high-density regions, to guide the optimization of low-evidence predictions that may appear near decision boundaries. Furthermore, the diverse sizes and locations of lesions in ultrasound images pose a challenge for CNNs with local receptive fields, as they struggle to model global information. Therefore, we introduce Visual Mamba based on structured state space sequence models, which achieves long-range dependency with linear computational complexity, and we construct a novel hybrid CNN-Mamba framework. During training, the collaboration between the CNN branch and the Mamba branch in the proposed framework draws inspiration from each other based on the EGC strategy. Experiments demonstrate the competitiveness of the proposed method. Dataset and code will be available on https://github.com/GtLinyer/MambaEviScrib.

Jinhao Chai, Shiyi Mu, Shugong Xu

Understanding other drivers' intentions is crucial for safe driving. The role of taillights in conveying these intentions is underemphasized in current autonomous driving systems. Accurately identifying taillight signals is essential for predicting vehicle behavior and preventing collisions. Open-source taillight datasets are scarce, often small and inconsistently annotated. To address this gap, we introduce a new large-scale taillight dataset called TLD. Sourced globally, our dataset covers diverse traffic scenarios. To our knowledge, TLD is the first dataset to separately annotate brake lights and turn signals in real driving scenarios. We collected 17.78 hours of driving videos from the internet. This dataset consists of 152k labeled image frames sampled at a rate of 2 Hz, along with 1.5 million unlabeled frames interspersed throughout. Additionally, we have developed a two-stage vehicle light detection model consisting of two primary modules: a vehicle detector and a taillight classifier. Initially, YOLOv10 and DeepSORT captured consecutive vehicle images over time. Subsequently, the two classifiers work simultaneously to determine the states of the brake lights and turn signals. A post-processing procedure is then used to eliminate noise caused by misidentifications and provide the taillight states of the vehicle within a given time frame. Our method shows exceptional performance on our dataset, establishing a benchmark for vehicle taillight detection. The dataset is available at https://huggingface.co/datasets/ChaiJohn/TLD/tree/main

Zhiqi Ai, Han Cheng, Shiyi Mu et al.

User-defined keyword spotting (KWS) is crucial for personalized voice interaction, yet existing methods face several challenges: (1) insufficient discriminability among confusable words, (2) performance inconsistency across speakers with varying pronunciations, and (3) high data cost to ensure reliable wake-word performance. In this paper, we introduce DMA-KWS, an efficient and robust framework for user-defined keyword spotting. First, it adopts a dual-stage matching pipeline: CTC decoding with streaming phoneme search to locate candidate segments, followed by QbyT with a phoneme matcher for fine-grained verification, enabling it to better distinguish confusable words. Next, multi-modal enrollment fuses user-specific speech with text embeddings to further improve accuracy for registered users. Finally, a parameter-efficient continual adaptation mechanism performs lightweight updates using synthetic and real data. Extensive experiments demonstrate the superior performance of DMA-KWS. On the LibriPhrase Hard subset, it achieves 97.85% AUC and 6.13% EER, reaching state-of-the-art performance. In speaker-dependent settings, DMA-KWS consistently outperforms text-only enrollment, demonstrating significant performance gains. Moreover, the proposed parameter-efficient fine-tuning mechanism adapts DMA-KWS with only 187k updated parameters, further enhancing KWS performance while ensuring suitability for on-device deployment.

Anqi Liu, Shiyi Mu, Shugong Xu

Autonomous driving algorithms usually employ sRGB images as model input due to their compatibility with the human visual system. However, visually pleasing sRGB images are possibly sub-optimal for downstream tasks when compared to RAW images. The availability of RAW images is constrained by the difficulties in collecting real-world driving data and the associated challenges of annotation. To address this limitation and support research in RAW-domain driving perception, we design a novel and ultra-lightweight RAW reconstruction method. The proposed model introduces a learnable color correction matrix (CCM), which uses only a single convolutional layer to approximate the complex inverse image signal processor (ISP). Experimental results demonstrate that simulated RAW (simRAW) images generated by our method provide performance improvements equivalent to those produced by more complex inverse ISP methods when pretraining RAW-domain object detectors, which highlights the effectiveness and practicality of our approach.

Kangcheng Zhou, Jun Jiang, Qing Zhang et al.

Interpretability is significant in computational pathology, leading to the development of multimodal information integration from histopathological image and corresponding text data.However, existing multimodal methods have limited interpretability due to the lack of high-quality dataset that support explicit reasoning and inference and simple reasoning process.To address the above problems, we introduce a novel multimodal pathology large language model with strong reasoning capabilities.To improve the generation of accurate and contextually relevant textual descriptions, we design a semantic reward strategy integrated with group relative policy optimization.We construct a high-quality pathology visual question answering (VQA) dataset, specifically designed to support complex reasoning tasks.Comprehensive experiments conducted on this dataset demonstrate that our method outperforms state-of-the-art methods, even when trained with only 20% of the data.Our method also achieves comparable performance on downstream zero-shot image classification task compared with CLIP.

Kongwu Huang, Shiyi Mu, Jun Jiang et al.

Scaling laws have achieved success in LLM and foundation models. To explore their potential in ISAC research, we propose Great-X. This single-engine multimodal data twin platform reconstructs the ray-tracing computation of Sionna within Unreal Engine and is deeply integrated with autonomous driving tools. This enables efficient and synchronized simulation of multimodal data, including CSI, RGB, Radar, and LiDAR. Based on this platform, we construct an open-source, large-scale, low-altitude UAV multimodal synaesthesia dataset named Great-MSD, and propose a baseline CSI-based UAV 3D localization algorithm, demonstrating its feasibility and generalizability across different CSI simulation engines. The related code and dataset will be made available at: https://github.com/hkw-xg/Great-MCD.

Yilin Gao, Shuguang Dou, Junzhou Li et al.

Recent advances in diffusion models have achieved remarkable success in isolated computer vision tasks such as text-to-image generation, depth estimation, and optical flow. However, these models are often restricted by a ``single-task-single-model'' paradigm, severely limiting their generalizability and scalability in multi-task scenarios. Motivated by the cross-domain generalization ability of large language models, we propose a universal visual perception framework based on flow matching that can generate diverse visual representations across multiple tasks. Our approach formulates the process as a universal flow-matching problem from image patch tokens to task-specific representations rather than an independent generation or regression problem. By leveraging a strong self-supervised foundation model as the anchor and introducing a multi-scale, circular task embedding mechanism, our method learns a universal velocity field to bridge the gap between heterogeneous tasks, supporting efficient and flexible representation transfer. Extensive experiments on classification, detection, segmentation, depth estimation, and image-text retrieval demonstrate that our model achieves competitive performance in both zero-shot and fine-tuned settings, outperforming prior generalist and several specialist models. Ablation studies further validate the robustness, scalability, and generalization of our framework. Our work marks a significant step towards general-purpose visual perception, providing a solid foundation for future research in universal vision modeling.

Shiyi Mu, Zichong Gu, Zhiqi Ai et al.

Compared to monocular 3D object detection, stereo-based 3D methods offer significantly higher accuracy but still suffer from high computational overhead and latency. The state-of-the-art stereo 3D detection method achieves twice the accuracy of monocular approaches, yet its inference speed is only half as fast. In this paper, we propose StereoDETR, an efficient stereo 3D object detection framework based on DETR. StereoDETR consists of two branches: a monocular DETR branch and a stereo branch. The DETR branch is built upon 2D DETR with additional channels for predicting object scale, orientation, and sampling points. The stereo branch leverages low-cost multi-scale disparity features to predict object-level depth maps. These two branches are coupled solely through a differentiable depth sampling strategy. To handle occlusion, we introduce a constrained supervision strategy for sampling points without requiring extra annotations. StereoDETR achieves real-time inference and is the first stereo-based method to surpass monocular approaches in speed. It also achieves competitive accuracy on the public KITTI benchmark, setting new state-of-the-art results on pedestrian and cyclist subsets. The code is available at https://github.com/shiyi-mu/StereoDETR-OPEN.

Shiyi Mu, Zichong Gu, Hanqi Lyu et al.

3D detection technology is widely used in the field of autonomous driving, with its application scenarios gradually expanding from enclosed highways to open conventional roads. For rare anomaly categories that appear on the road, 3D detection models trained on closed sets often misdetect or fail to detect anomaly objects. To address this risk, it is necessary to enhance the generalization ability of 3D detection models for targets of arbitrary shapes and to possess the capability to filter out anomalies. The generalization of 3D detection is limited by two factors: the coupled training of 2D and 3D, and the insufficient diversity in the scale distribution of training samples. This paper proposes a Stereo-based 3D Anomaly object Detection (S3AD) algorithm, which decouples the training strategy of 3D and 2D to release the generalization ability for arbitrary 3D foreground detection, and proposes an anomaly scoring algorithm based on foreground confidence prediction, achieving target-level anomaly scoring. In order to further verify and enhance the generalization of anomaly detection, we use a 3D rendering method to synthesize two augmented reality binocular stereo 3D detection datasets which named KITTI-AR. KITTI-AR extends upon KITTI by adding 97 new categories, totaling 6k pairs of stereo images. The KITTI-AR-ExD subset includes 39 common categories as extra training data to address the sparse sample distribution issue. Additionally, 58 rare categories form the KITTI-AR-OoD subset, which are not used in training to simulate zero-shot scenarios in real-world settings, solely for evaluating 3D anomaly detection. Finally, the performance of the algorithm and the dataset is verified in the experiments. (Code and dataset can be obtained at https://github.com/shiyi-mu/S3AD-Code).

Jun Jiang, Wenjun Yu, Yunfan Li et al.

In the field of artificial intelligence, self-supervised learning has demonstrated superior generalization capabilities by leveraging large-scale unlabeled datasets for pretraining, which is especially critical for wireless communication models to adapt to a variety of scenarios. This paper innovatively treats Channel State Information (CSI) and Channel Impulse Response (CIR) as naturally aligned multi-modal data and proposes the first MIMO wireless channel foundation model, named CSI-CLIP. By effectively capturing the joint representations of both CIR and CSI, CSI-CLIP exhibits remarkable adaptability across scenarios and robust feature extraction capabilities. Experimental results show that in positioning task, CSI-CLIP reduces the mean error distance by 22%; in beam management task, it increases accuracy by 1% compared to traditional supervised methods, as well as in the channel identification task. These improvements not only highlight the potential and value of CSI-CLIP in integrating sensing and communication but also demonstrate its significant advantages over existing techniques. Moreover, viewing CSI and CIR as multi-modal pairs and contrastive learning for wireless channel foundation model open up new research directions in the domain of MIMO wireless communications.

Guangjin Pan, Yuan Gao, Yilin Gao et al.

Wireless positioning technologies hold significant value for applications in autonomous driving, extended reality (XR), unmanned aerial vehicles (UAVs), and more. With the advancement of artificial intelligence (AI), leveraging AI to enhance positioning accuracy and robustness has emerged as a field full of potential. Driven by the requirements and functionalities defined in the 3rd Generation Partnership Project (3GPP) standards, AI/machine learning (ML)-based cellular positioning is becoming a key technology to overcome the limitations of traditional methods. This paper presents a comprehensive survey of AI-driven cellular positioning. We begin by reviewing the fundamentals of wireless positioning and AI models, analyzing their respective challenges and synergies. We provide a comprehensive review of the evolution of 3GPP positioning standards, with a focus on the integration of AI/ML in current and upcoming standard releases. Guided by the 3GPP-defined taxonomy, we categorize and summarize state-of-the-art (SOTA) research into two major classes: AI/ML-assisted positioning and direct AI/ML-based positioning. The former includes line-of-sight (LOS)/non-line-of-sight (NLOS) detection, time of arrival (TOA)/time difference of arrival (TDOA) estimation, and angle prediction; the latter encompasses fingerprinting, knowledge-assisted learning, and channel charting. Furthermore, we review representative public datasets and conduct performance evaluations of AI-based positioning algorithms using these datasets. Finally, we conclude by summarizing the challenges and opportunities of AI-driven wireless positioning.

Yanliang Jin, Yifan Wu, Yuan Gao et al.

The emergence of 6th generation (6G) mobile networks brings new challenges in supporting high-mobility communications, particularly in addressing the issue of channel aging. While existing channel prediction methods offer improved accuracy at the expense of increased computational complexity, limiting their practical application in mobile networks. To address these challenges, we present LinFormer, an innovative channel prediction framework based on a scalable, all-linear, encoder-only Transformer model. Our approach, inspired by natural language processing (NLP) models such as BERT, adapts an encoder-only architecture specifically for channel prediction tasks. We propose replacing the computationally intensive attention mechanism commonly used in Transformers with a time-aware multi-layer perceptron (TMLP), significantly reducing computational demands. The inherent time awareness of TMLP module makes it particularly suitable for channel prediction tasks. We enhance LinFormer's training process by employing a weighted mean squared error loss (WMSELoss) function and data augmentation techniques, leveraging larger, readily available communication datasets. Our approach achieves a substantial reduction in computational complexity while maintaining high prediction accuracy, making it more suitable for deployment in cost-effective base stations (BS). Comprehensive experiments using both simulated and measured data demonstrate that LinFormer outperforms existing methods across various mobility scenarios, offering a promising solution for future wireless communication systems.

Anqing Jiang, Jinhao Chai, Yu Gao et al.

Recent advancements in high-definition \emph{HD} map construction have demonstrated the effectiveness of dense representations, which heavily rely on computationally intensive bird's-eye view \emph{BEV} features. While sparse representations offer a more efficient alternative by avoiding dense BEV processing, existing methods often lag behind due to the lack of tailored designs. These limitations have hindered the competitiveness of sparse representations in online HD map construction. In this work, we systematically revisit and enhance sparse representation techniques, identifying key architectural and algorithmic improvements that bridge the gap with--and ultimately surpass--dense approaches. We introduce a dedicated network architecture optimized for sparse map feature extraction, a sparse-dense segmentation auxiliary task to better leverage geometric and semantic cues, and a denoising module guided by physical priors to refine predictions. Through these enhancements, our method achieves state-of-the-art performance on the nuScenes dataset, significantly advancing HD map construction and centerline detection. Specifically, SparseMeXt-Tiny reaches a mean average precision \emph{mAP} of 55.5% at 32 frames per second \emph{fps}, while SparseMeXt-Base attains 65.2% mAP. Scaling the backbone and decoder further, SparseMeXt-Large achieves an mAP of 68.9% at over 20 fps, establishing a new benchmark for sparse representations in HD map construction. These results underscore the untapped potential of sparse methods, challenging the conventional reliance on dense representations and redefining efficiency-performance trade-offs in the field.

Xiaojing Chen, Haiqi Yu, Wei Ni et al.

The rapid emergence of Large Language Models (LLMs) has catalyzed Agentic artificial intelligence (AI), autonomous systems integrating perception, reasoning, and action into closed-loop pipelines for continuous adaptation. While unlocking transformative applications in mobile edge computing, autonomous systems, and next-generation wireless networks, this paradigm creates fundamental energy challenges through iterative inference and persistent data exchange. Unlike traditional AI where bottlenecks are computational Floating Point Operations (FLOPs), Agentic AI faces compounding computational and communication energy costs. In this survey, we propose an energy accounting framework identifying computational and communication costs across the Perception-Reasoning-Action cycle. We establish a unified taxonomy spanning model simplification, computation control, input and attention optimization, and hardware-aware inference. We explore cross-layer co-design strategies jointly optimizing model parameters, wireless transmissions, and edge resources. Finally, we identify open challenges of federated green learning, carbon-aware agency, 6th generation mobile communication (6G)-native Agentic AI, and self-sustaining systems, providing a roadmap for scalable autonomous intelligence.

Jinhao Chai, Anqing Jiang, Hao Jiang et al.

End-to-end multi-modal planning has become a transformative paradigm in autonomous driving, effectively addressing behavioral multi-modality and the generalization challenge in long-tail scenarios. We propose AnchDrive, a framework for end-to-end driving that effectively bootstraps a diffusion policy to mitigate the high computational cost of traditional generative models. Rather than denoising from pure noise, AnchDrive initializes its planner with a rich set of hybrid trajectory anchors. These anchors are derived from two complementary sources: a static vocabulary of general driving priors and a set of dynamic, context-aware trajectories. The dynamic trajectories are decoded in real-time by a Transformer that processes dense and sparse perceptual features. The diffusion model then learns to refine these anchors by predicting a distribution of trajectory offsets, enabling fine-grained refinement. This anchor-based bootstrapping design allows for efficient generation of diverse, high-quality trajectories. Experiments on the NAVSIM benchmark confirm that AnchDrive sets a new state-of-the-art and shows strong generalizability

Yilin Gao, Kangyi Chen, Zhongxing Peng et al.

Current visual foundation models (VFMs) face a fundamental limitation in transferring knowledge from vision language models (VLMs), while VLMs excel at modeling cross-modal interactions through unified representation spaces, existing VFMs predominantly adopt result-oriented paradigms that neglect the underlying interaction processes. This representational discrepancy hinders effective knowledge transfer and limits generalization across diverse vision tasks. We propose Learning from Interactions (LFI), a cognitive-inspired framework that addresses this gap by explicitly modeling visual understanding as an interactive process. Our key insight is that capturing the dynamic interaction patterns encoded in pre-trained VLMs enables more faithful and efficient knowledge transfer to VFMs. The approach centers on two technical innovations, Interaction Queries, which maintain persistent relational structures across network layers, and interaction-based supervision, derived from the cross-modal attention mechanisms of VLMs. Comprehensive experiments demonstrate consistent improvements across multiple benchmarks, achieving 3.3 and 1.6mAP/2.4AP absolute gains on TinyImageNet classification and COCO detection/segmentation respectively, with minimal parameter overhead and faster convergence. The framework particularly excels in cross-domain settings, delivering 2.4 and 9.3 zero-shot improvements on PACS and VLCS. Human evaluations further confirm its cognitive alignment, outperforming result-oriented methods by 2.7 times in semantic consistency metrics.

Bo Wu, Zhiqi Ai, Jun Jiang et al.

Label ambiguity poses a significant challenge in age estimation tasks. Most existing methods address this issue by modeling correlations between adjacent age groups through label distribution learning. However, they often overlook the varying degrees of ambiguity present across different age stages. In this paper, we propose a Stage-wise Adaptive Label Distribution Learning (SA-LDL) algorithm, which leverages the observation -- revealed through our analysis of embedding similarities between an anchor and all other ages -- that label ambiguity exhibits clear stage-wise patterns. By jointly employing stage-wise adaptive variance modeling and weighted loss function, SA-LDL effectively captures the complex and structured nature of label ambiguity, leading to more accurate and robust age estimation. Extensive experiments demonstrate that SA-LDL achieves competitive performance, with MAE of 1.74 and 2.15 on the MORPH-II and FG-NET datasets.

Zhiqi Ai, Meixuan Bao, Zhiyong Chen et al.

The performance of speaker verification systems is adversely affected by speaker aging. However, due to challenges in data collection, particularly the lack of sustained and large-scale longitudinal data for individuals, research on speaker aging remains difficult. In this paper, we present VoxAging, a large-scale longitudinal dataset collected from 293 speakers (226 English speakers and 67 Mandarin speakers) over several years, with the longest time span reaching 17 years (approximately 900 weeks). For each speaker, the data were recorded at weekly intervals. We studied the phenomenon of speaker aging and its effects on advanced speaker verification systems, analyzed individual speaker aging processes, and explored the impact of factors such as age group and gender on speaker aging research.

Tao Wen, Jiepeng Wang, Yabo Chen et al.

Accurate and generalizable metric depth estimation is crucial for various computer vision applications but remains challenging due to the diverse depth scales encountered in indoor and outdoor environments. In this paper, we introduce Metric-Solver, a novel sliding anchor-based metric depth estimation method that dynamically adapts to varying scene scales. Our approach leverages an anchor-based representation, where a reference depth serves as an anchor to separate and normalize the scene depth into two components: scaled near-field depth and tapered far-field depth. The anchor acts as a normalization factor, enabling the near-field depth to be normalized within a consistent range while mapping far-field depth smoothly toward zero. Through this approach, any depth from zero to infinity in the scene can be represented within a unified representation, effectively eliminating the need to manually account for scene scale variations. More importantly, for the same scene, the anchor can slide along the depth axis, dynamically adjusting to different depth scales. A smaller anchor provides higher resolution in the near-field, improving depth precision for closer objects while a larger anchor improves depth estimation in far regions. This adaptability enables the model to handle depth predictions at varying distances and ensure strong generalization across datasets. Our design enables a unified and adaptive depth representation across diverse environments. Extensive experiments demonstrate that Metric-Solver outperforms existing methods in both accuracy and cross-dataset generalization.

Yanzan Sun, Jiacheng Qiu, Guangjin Pan et al.

Extended reality (XR), blending virtual and real worlds, is a key application of future networks. While AI advancements enhance XR capabilities, they also impose significant computational and energy challenges on lightweight XR devices. In this paper, we developed a distributed queue model for multi-task DNN inference, addressing issues of resource competition and queue coupling. In response to the challenges posed by the high energy consumption and limited resources of XR devices, we designed a dual time-scale joint optimization strategy for model partitioning and resource allocation, formulated as a bi-level optimization problem. This strategy aims to minimize the total energy consumption of XR devices while ensuring queue stability and adhering to computational and communication resource constraints. To tackle this problem, we devised a Lyapunov-guided Proximal Policy Optimization algorithm, named LyaPPO. Numerical results demonstrate that the LyaPPO algorithm outperforms the baselines, achieving energy conservation of 24.79% to 46.14% under varying resource capacities. Specifically, the proposed algorithm reduces the energy consumption of XR devices by 24.29% to 56.62% compared to baseline algorithms.

Yifei Jin, Ali Maatouk, Sarunas Girdzijauskas et al.

Wireless ray-tracing (RT) is emerging as a key tool for three-dimensional (3D) wireless channel modeling, driven by advances in graphical rendering. Current approaches struggle to accurately model beyond 5G (B5G) network signaling, which often operates at higher frequencies and is more susceptible to environmental conditions and changes. Existing online learning solutions require real-time environmental supervision during training, which is both costly and incompatible with GPU-based processing. In response, we propose a novel approach that redefines ray trajectory generation as a sequential decision-making problem, leveraging generative models to jointly learn the optical, physical, and signal properties within each designated environment. Our work introduces the Scene-Aware Neural Decision Wireless Channel Raytracing Hierarchy (SANDWICH), an innovative offline, fully differentiable approach that can be trained entirely on GPUs. SANDWICH offers superior performance compared to existing online learning methods, outperforms the baseline by 4e^-2 radian in RT accuracy, and only fades 0.5 dB away from toplined channel gain estimation.

Xiaojing Chen, Zhenyuan Li, Wei Ni et al.

Federated learning (FL) is a viable technique to train a shared machine learning model without sharing data. Hierarchical FL (HFL) system has yet to be studied regrading its multiple levels of energy, computation, communication, and client scheduling, especially when it comes to clients relying on energy harvesting to power their operations. This paper presents a new two-phase deep deterministic policy gradient (DDPG) framework, referred to as ``TP-DDPG'', to balance online the learning delay and model accuracy of an FL process in an energy harvesting-powered HFL system. The key idea is that we divide optimization decisions into two groups, and employ DDPG to learn one group in the first phase, while interpreting the other group as part of the environment to provide rewards for training the DDPG in the second phase. Specifically, the DDPG learns the selection of participating clients, and their CPU configurations and the transmission powers. A new straggler-aware client association and bandwidth allocation (SCABA) algorithm efficiently optimizes the other decisions and evaluates the reward for the DDPG. Experiments demonstrate that with substantially reduced number of learnable parameters, the TP-DDPG can quickly converge to effective polices that can shorten the training time of HFL by 39.4% compared to its benchmarks, when the required test accuracy of HFL is 0.9.

Zhiqi Ai, Zhiyong Chen, Shugong Xu

In this paper, we propose MM-KWS, a novel approach to user-defined keyword spotting leveraging multi-modal enrollments of text and speech templates. Unlike previous methods that focus solely on either text or speech features, MM-KWS extracts phoneme, text, and speech embeddings from both modalities. These embeddings are then compared with the query speech embedding to detect the target keywords. To ensure the applicability of MM-KWS across diverse languages, we utilize a feature extractor incorporating several multilingual pre-trained models. Subsequently, we validate its effectiveness on Mandarin and English tasks. In addition, we have integrated advanced data augmentation tools for hard case mining to enhance MM-KWS in distinguishing confusable words. Experimental results on the LibriPhrase and WenetPhrase datasets demonstrate that MM-KWS outperforms prior methods significantly.

Yue Tao, Zhiwei Jia, Runze Ma et al.

Scene text recognition (STR) is an important bridge between images and text, attracting abundant research attention. While convolutional neural networks (CNNS) have achieved remarkable progress in this task, most of the existing works need an extra module (context modeling module) to help CNN to capture global dependencies to solve the inductive bias and strengthen the relationship between text features. Recently, the transformer has been proposed as a promising network for global context modeling by self-attention mechanism, but one of the main shortcomings, when applied to recognition, is the efficiency. We propose a 1-D split to address the challenges of complexity and replace the CNN with the transformer encoder to reduce the need for a context modeling module. Furthermore, recent methods use a frozen initial embedding to guide the decoder to decode the features to text, leading to a loss of accuracy. We propose to use a learnable initial embedding learned from the transformer encoder to make it adaptive to different input images. Above all, we introduce a novel architecture for text recognition, named TRansformer-based text recognizer with Initial embedding Guidance (TRIG), composed of three stages (transformation, feature extraction, and prediction). Extensive experiments show that our approach can achieve state-of-the-art on text recognition benchmarks.

Dinghao Fan, Hengjie Lu, Shugong Xu et al.

Gesture recognition is getting more and more popular due to various application possibilities in human-machine interaction. Existing multi-modal gesture recognition systems take multi-modal data as input to improve accuracy, but such methods require more modality sensors, which will greatly limit their application scenarios. Therefore we propose an end-to-end multi-task learning framework in training 2D convolutional neural networks. The framework can use the depth modality to improve accuracy during training and save costs by using only RGB modality during inference. Our framework is trained to learn a representation for multi-task learning: gesture segmentation and gesture recognition. Depth modality contains the prior information for the location of the gesture. Therefore it can be used as the supervision for gesture segmentation. A plug-and-play module named Multi-Scale-Decoder is designed to realize gesture segmentation, which contains two sub-decoder. It is used in the lower stage and higher stage respectively, and can help the network pay attention to key target areas, ignore irrelevant information, and extract more discriminant features. Additionally, the MSD module and depth modality are only used in the training stage to improve gesture recognition performance. Only RGB modality and network without MSD are required during inference. Experimental results on three public gesture recognition datasets show that our proposed method provides superior performance compared with existing gesture recognition frameworks. Moreover, using the proposed plug-and-play MSD in other 2D CNN-based frameworks also get an excellent accuracy improvement.

Hang Cheng, Shugong Xu, Xiufeng Jiang et al.

Image matting is an important computer vision problem. Many existing matting methods require a hand-made trimap to provide auxiliary information, which is very expensive and limits the real world usage. Recently, some trimap-free methods have been proposed, which completely get rid of any user input. However, their performance lag far behind trimap-based methods due to the lack of guidance information. In this paper, we propose a matting method that use Flexible Guidance Input as user hint, which means our method can use trimap, scribblemap or clickmap as guidance information or even work without any guidance input. To achieve this, we propose Progressive Trimap Deformation(PTD) scheme that gradually shrink the area of the foreground and background of the trimap with the training step increases and finally become a scribblemap. To make our network robust to any user scribble and click, we randomly sample points on foreground and background and perform curve fitting. Moreover, we propose Semantic Fusion Module(SFM) which utilize the Feature Pyramid Enhancement Module(FPEM) and Joint Pyramid Upsampling(JPU) in matting task for the first time. The experiments show that our method can achieve state-of-the-art results comparing with existing trimap-based and trimap-free methods.

Jingchao Chen, Shiyi Mu, Shugong Xu et al.

Although lots of progress were made in Text Recognition/OCR in recent years, the task of font recognition is remaining challenging. The main challenge lies in the subtle difference between these similar fonts, which is hard to distinguish. This paper proposes a novel font recognizer with a pluggable module solving the font recognition task. The pluggable module hides the most discriminative accessible features and forces the network to consider other complicated features to solve the hard examples of similar fonts, called HE Block. Compared with the available public font recognition systems, our proposed method does not require any interactions at the inference stage. Extensive experiments demonstrate that HENet achieves encouraging performance, including on character-level dataset Explor_all and word-level dataset AdobeVFR

Zhiwei Jia, Shugong Xu, Shiyi Mu et al.

Although recent works based on deep learning have made progress in improving recognition accuracy on scene text recognition, how to handle low-quality text images in end-to-end deep networks remains a research challenge. In this paper, we propose an Iterative Fusion based Recognizer (IFR) for low quality scene text recognition, taking advantage of refined text images input and robust feature representation. IFR contains two branches which focus on scene text recognition and low quality scene text image recovery respectively. We utilize an iterative collaboration between two branches, which can effectively alleviate the impact of low quality input. A feature fusion module is proposed to strengthen the feature representation of the two branches, where the features from the Recognizer are Fused with image Restoration branch, referred to as RRF. Without changing the recognition network structure, extensive quantitative and qualitative experimental results show that the proposed method significantly outperforms the baseline methods in boosting the recognition accuracy of benchmark datasets and low resolution images in TextZoom dataset.

Youxuan Ma, Zongze Ren, Shugong Xu

In recent years, synthetic speech generated by advanced text-to-speech (TTS) and voice conversion (VC) systems has caused great harms to automatic speaker verification (ASV) systems, urging us to design a synthetic speech detection system to protect ASV systems. In this paper, we propose a new speech anti-spoofing model named ResWavegram-Resnet (RW-Resnet). The model contains two parts, Conv1D Resblocks and backbone Resnet34. The Conv1D Resblock is based on the Conv1D block with a residual connection. For the first part, we use the raw waveform as input and feed it to the stacked Conv1D Resblocks to get the ResWavegram. Compared with traditional methods, ResWavegram keeps all the information from the audio signal and has a stronger ability in extracting features. For the second part, the extracted features are fed to the backbone Resnet34 for the spoofed or bonafide decision. The ASVspoof2019 logical access (LA) corpus is used to evaluate our proposed RW-Resnet. Experimental results show that the RW-Resnet achieves better performance than other state-of-the-art anti-spoofing models, which illustrates its effectiveness in detecting synthetic speech attacks.

Hengjie Lu, Shugong Xu, Shan Cao

Depth completion aims to generate a dense depth map from the sparse depth map and aligned RGB image. However, current depth completion methods use extremely expensive 64-line LiDAR(about $100,000) to obtain sparse depth maps, which will limit their application scenarios. Compared with the 64-line LiDAR, the single-line LiDAR is much less expensive and much more robust. Therefore, we propose a method to tackle the problem of single-line depth completion, in which we aim to generate a dense depth map from the single-line LiDAR info and the aligned RGB image. A single-line depth completion dataset is proposed based on the existing 64-line depth completion dataset(KITTI). A network called Semantic Guided Two-Branch Network(SGTBN) which contains global and local branches to extract and fuse global and local info is proposed for this task. A Semantic guided depth upsampling module is used in our network to make full use of the semantic info in RGB images. Except for the usual MSE loss, we add the virtual normal loss to increase the constraint of high-order 3D geometry in our network. Our network outperforms the state-of-the-art in the single-line depth completion task. Besides, compared with the monocular depth estimation, our method also has significant advantages in precision and model size.

Xiufeng Jiang, Shugong Xu, Shunqing Zhang et al.

Text detection/localization, as an important task in computer vision, has witnessed substantialadvancements in methodology and performance with convolutional neural networks. However, the vastmajority of popular methods use rectangles or quadrangles to describe text regions. These representationshave inherent drawbacks, especially relating to dense adjacent text and loose regional text boundaries,which usually cause difficulty detecting arbitrarily shaped text. In this paper, we propose a novel text regionrepresentation method, with a robust pipeline, which can precisely detect dense adjacent text instances witharbitrary shapes. We consider a text instance to be composed of an adaptive central text region mask anda corresponding expanding ratio between the central text region and the full text region. More specifically,our pipeline generates adaptive central text regions and corresponding expanding ratios with a proposedtraining strategy, followed by a new proposed post-processing algorithm which expands central text regionsto the complete text instance with the corresponding expanding ratios. We demonstrated that our new textregion representation is effective, and that the pipeline can precisely detect closely adjacent text instances ofarbitrary shapes. Experimental results on common datasets demonstrate superior performance o

Jiajun Zhu, Xiufeng Jiang, Zhiwei Jia et al.

Recently, video scene text detection has received increasing attention due to its comprehensive applications. However, the lack of annotated scene text video datasets has become one of the most important problems, which hinders the development of video scene text detection. The existing scene text video datasets are not large-scale due to the expensive cost caused by manual labeling. In addition, the text instances in these datasets are too clear to be a challenge. To address the above issues, we propose a tracking based semi-automatic labeling strategy for scene text videos in this paper. We get semi-automatic scene text annotation by labeling manually for the first frame and tracking automatically for the subsequent frames, which avoid the huge cost of manual labeling. Moreover, a paired low-quality scene text video dataset named Text-RBL is proposed, consisting of raw videos, blurry videos, and low-resolution videos, labeled by the proposed convenient semi-automatic labeling strategy. Through an averaging operation and bicubic down-sampling operation over the raw videos, we can efficiently obtain blurry videos and low-resolution videos paired with raw videos separately. To verify the effectiveness of Text-RBL, we propose a baseline model combined with the text detector and tracker for video scene text detection. Moreover, a failure detection scheme is designed to alleviate the baseline model drift issue caused by complex scenes. Extensive experiments demonstrate that Text-RBL with paired low-quality videos labeled by the semi-automatic method can significantly improve the performance of the text detector in low-quality scenes.

Xudong Chen, Shugong Xu, Qiaobin Ji et al.

Face Anti-spoofing (FAS) is a challenging problem due to complex serving scenarios and diverse face presentation attack patterns. Especially when captured images are low-resolution, blurry, and coming from different domains, the performance of FAS will degrade significantly. The existing multi-modal FAS datasets rarely pay attention to the cross-domain problems under deployment scenarios, which is not conducive to the study of model performance. To solve these problems, we explore the fine-grained differences between multi-modal cameras and construct a cross-domain multi-modal FAS dataset under surveillance scenarios called GREAT-FASD-S. Besides, we propose an Attention based Face Anti-spoofing network with Feature Augment (AFA) to solve the FAS towards low-quality face images. It consists of the depthwise separable attention module (DAM) and the multi-modal based feature augment module (MFAM). Our model can achieve state-of-the-art performance on the CASIA-SURF dataset and our proposed GREAT-FASD-S dataset.

Peiyao Wang, Weixin Luo, Yanyu Xu et al.

Spatial Description Resolution, as a language-guided localization task, is proposed for target location in a panoramic street view, given corresponding language descriptions. Explicitly characterizing an object-level relationship while distilling spatial relationships are currently absent but crucial to this task. Mimicking humans, who sequentially traverse spatial relationship words and objects with a first-person view to locate their target, we propose a novel spatial relationship induced (SIRI) network. Specifically, visual features are firstly correlated at an implicit object-level in a projected latent space; then they are distilled by each spatial relationship word, resulting in each differently activated feature representing each spatial relationship. Further, we introduce global position priors to fix the absence of positional information, which may result in global positional reasoning ambiguities. Both the linguistic and visual features are concatenated to finalize the target localization. Experimental results on the Touchdown show that our method is around 24\% better than the state-of-the-art method in terms of accuracy, measured by an 80-pixel radius. Our method also generalizes well on our proposed extended dataset collected using the same settings as Touchdown.

Yuxi Li, Weiyao Lin, Tao Wang et al.

The task of spatial-temporal action detection has attracted increasing attention among researchers. Existing dominant methods solve this problem by relying on short-term information and dense serial-wise detection on each individual frames or clips. Despite their effectiveness, these methods showed inadequate use of long-term information and are prone to inefficiency. In this paper, we propose for the first time, an efficient framework that generates action tube proposals from video streams with a single forward pass in a sparse-to-dense manner. There are two key characteristics in this framework: (1) Both long-term and short-term sampled information are explicitly utilized in our spatiotemporal network, (2) A new dynamic feature sampling module (DTS) is designed to effectively approximate the tube output while keeping the system tractable. We evaluate the efficacy of our model on the UCF101-24, JHMDB-21 and UCFSports benchmark datasets, achieving promising results that are competitive to state-of-the-art methods. The proposed sparse-to-dense strategy rendered our framework about 7.6 times more efficient than the nearest competitor.

Yuxi Li, Weiyao Lin, John See et al.

Most current pipelines for spatio-temporal action localization connect frame-wise or clip-wise detection results to generate action proposals, where only local information is exploited and the efficiency is hindered by dense per-frame localization. In this paper, we propose Coarse-to-Fine Action Detector (CFAD),an original end-to-end trainable framework for efficient spatio-temporal action localization. The CFAD introduces a new paradigm that first estimates coarse spatio-temporal action tubes from video streams, and then refines the tubes' location based on key timestamps. This concept is implemented by two key components, the Coarse and Refine Modules in our framework. The parameterized modeling of long temporal information in the Coarse Module helps obtain accurate initial tube estimation, while the Refine Module selectively adjusts the tube location under the guidance of key timestamps. Against other methods, theproposed CFAD achieves competitive results on action detection benchmarks of UCF101-24, UCFSports and JHMDB-21 with inference speed that is 3.3x faster than the nearest competitors.

Zhichao Zhang, Shugong Xu, Shunqing Zhang et al.

Environmental sound classification (ESC) is a challenging problem due to the complexity of sounds. The classification performance is heavily dependent on the effectiveness of representative features extracted from the environmental sounds. However, ESC often suffers from the semantically irrelevant frames and silent frames. In order to deal with this, we employ a frame-level attention model to focus on the semantically relevant frames and salient frames. Specifically, we first propose a convolutional recurrent neural network to learn spectro-temporal features and temporal correlations. Then, we extend our convolutional RNN model with a frame-level attention mechanism to learn discriminative feature representations for ESC. We investigated the classification performance when using different attention scaling function and applying different layers. Experiments were conducted on ESC-50 and ESC-10 datasets. Experimental results demonstrated the effectiveness of the proposed method and our method achieved the state-of-the-art or competitive classification accuracy with lower computational complexity. We also visualized our attention results and observed that the proposed attention mechanism was able to lead the network tofocus on the semantically relevant parts of environmental sounds.

Yan Liu, Zhiyuan Jiang, Shunqing Zhang et al.

Ultra-Reliable and Low-Latency Communications (URLLC) services in vehicular networks on millimeter-wave bands present a significant challenge, considering the necessity of constantly adjusting the beam directions. Conventional methods are mostly based on classical control theory, e.g., Kalman filter and its variations, which mainly deal with stationary scenarios. Therefore, severe application limitations exist, especially with complicated, dynamic Vehicle-to-Everything (V2X) channels. This paper gives a thorough study of this subject, by first modifying the classical approaches, e.g., Extended Kalman Filter (EKF) and Particle Filter (PF), for non-stationary scenarios, and then proposing a Reinforcement Learning (RL)-based approach that can achieve the URLLC requirements in a typical intersection scenario. Simulation results based on a commercial ray-tracing simulator show that enhanced EKF and PF methods achieve packet delay more than $10$ ms, whereas the proposed deep RL-based method can reduce the latency to about $6$ ms, by extracting context information from the training data.

Tianhao Qiao, Shunqing Zhang, Zhichao Zhang et al.

Environmental Sound Classification (ESC) is an important and challenging problem, and feature representation is a critical and even decisive factor in ESC. Feature representation ability directly affects the accuracy of sound classification. Therefore, the ESC performance is heavily dependent on the effectiveness of representative features extracted from the environmental sounds. In this paper, we propose a subspectrogram segmentation based ESC classification framework. In addition, we adopt the proposed Convolutional Recurrent Neural Network (CRNN) and score level fusion to jointly improve the classification accuracy. Extensive truncation schemes are evaluated to find the optimal number and the corresponding band ranges of sub-spectrograms. Based on the numerical experiments, the proposed framework can achieve 81.9% ESC classification accuracy on the public dataset ESC-50, which provides 9.1% accuracy improvement over traditional baseline schemes.

Zhiyong Chen, Zongze Ren, Shugong Xu

Learning a good speaker embedding is important for many automatic speaker recognition tasks, including verification, identification and diarization. The embeddings learned by softmax are not discriminative enough for open-set verification tasks. Angular based embedding learning target can achieve such discriminativeness by optimizing angular distance and adding margin penalty. We apply several different popular angular margin embedding learning strategies in this work and explicitly compare their performance on Voxceleb speaker recognition dataset. Observing the fact that encouraging inter-class separability is important when applying angular based embedding learning, we propose an exclusive inter-class regularization as a complement for angular based loss. We verify the effectiveness of these methods for learning a discriminative embedding space on ASV task with several experiments. These methods together, we manage to achieve an impressive result with 16.5% improvement on equal error rate (EER) and 18.2% improvement on minimum detection cost function comparing with baseline softmax systems.

Zongze Ren, Guofu Yang, Shugong Xu

In this paper, we present a two-stage language identification (LID) system based on a shallow ResNet14 followed by a simple 2-layer recurrent neural network (RNN) architecture, which was used for Xunfei (iFlyTek) Chinese Dialect Recognition Challenge and won the first place among 110 teams. The system trains an acoustic model (AM) firstly with connectionist temporal classification (CTC) to recognize the given phonetic sequence annotation and then train another RNN to classify dialect category by utilizing the intermediate features as inputs from the AM. Compared with a three-stage system we further explore, our results show that the two-stage system can achieve high accuracy for Chinese dialects recognition under both short utterance and long utterance conditions with less training time.

Zongze Ren, Zhiyong Chen, Shugong Xu

Speaker embeddings become growing popular in the text-independent speaker verification task. In this paper, we propose two improvements during the training stage. The improvements are both based on triplet cause the training stage and the evaluation stage of the baseline x-vector system focus on different aims. Firstly, we introduce triplet loss for optimizing the Euclidean distances between embeddings while minimizing the multi-class cross entropy loss. Secondly, we design an embedding similarity measurement network for controlling the similarity between the two selected embeddings. We further jointly train the two new methods with the original network and achieve state-of-the-art. The multi-task training synergies are shown with a 9% reduction equal error rate (EER) and detected cost function (DCF) on the 2016 NIST Speaker Recognition Evaluation (SRE) Test Set.

Zhichao Zhang, Shugong Xu, Tianhao Qiao et al.

Environmental sound classification (ESC) is a challenging problem due to the complexity of sounds. The ESC performance is heavily dependent on the effectiveness of representative features extracted from the environmental sounds. However, ESC often suffers from the semantically irrelevant frames and silent frames. In order to deal with this, we employ a frame-level attention model to focus on the semantically relevant frames and salient frames. Specifically, we first propose an convolutional recurrent neural network to learn spectro-temporal features and temporal correlations. Then, we extend our convolutional RNN model with a frame-level attention mechanism to learn discriminative feature representations for ESC. Experiments were conducted on ESC-50 and ESC-10 datasets. Experimental results demonstrated the effectiveness of the proposed method and achieved the state-of-the-art performance in terms of classification accuracy.

Xiaoyu Chen, Shugong Xu, Xudong Chen et al.

Transmission control protocol (TCP) congestion control is one of the key techniques to improve network performance. TCP congestion control algorithm identification (TCP identification) can be used to significantly improve network efficiency. Existing TCP identification methods can only be applied to limited number of TCP congestion control algorithms and focus on wired networks. In this paper, we proposed a machine learning based passive TCP identification method for wired and wireless networks. After comparing among three typical machine learning models, we concluded that the 4-layers Long Short Term Memory (LSTM) model achieves the best identification accuracy. Our approach achieves better than 98% accuracy in wired and wireless networks and works for newly proposed TCP congestion control algorithms.

Changhao Wu, Shugong Xu, Guocong Song et al.

Training a good deep learning model often requires a lot of annotated data. As a large amount of labeled data is typically difficult to collect and even more difficult to annotate, data augmentation and data generation are widely used in the process of training deep neural networks. However, there is no clear common understanding on how much labeled data is needed to get satisfactory performance. In this paper, we try to address such a question using vehicle license plate character recognition as an example application. We apply computer graphic scripts and Generative Adversarial Networks to generate and augment a large number of annotated, synthesized license plate images with realistic colors, fonts, and character composition from a small number of real, manually labeled license plate images. Generated and augmented data are mixed and used as training data for the license plate recognition network modified from DenseNet. The experimental results show that the model trained from the generated mixed training data has good generalization ability, and the proposed approach achieves a new state-of-the-art accuracy on Dataset-1 and AOLP, even with a very limited number of original real license plates. In addition, the accuracy improvement caused by data generation becomes more significant when the number of labeled images is reduced. Data augmentation also plays a more significant role when the number of labeled images is increased.

Zhichao Zhang, Shugong Xu, Shan Cao et al.

Environmental sound classification (ESC) is an important and challenging problem. In contrast to speech, sound events have noise-like nature and may be produced by a wide variety of sources. In this paper, we propose to use a novel deep convolutional neural network for ESC tasks. Our network architecture uses stacked convolutional and pooling layers to extract high-level feature representations from spectrogram-like features. Furthermore, we apply mixup to ESC tasks and explore its impacts on classification performance and feature distribution. Experiments were conducted on UrbanSound8K, ESC-50 and ESC-10 datasets. Our experimental results demonstrated that our ESC system has achieved the state-of-the-art performance (83.7%) on UrbanSound8K and competitive performance on ESC-50 and ESC-10.

Yuanzhouhan Cao, Tianqi Zhao, Ke Xian et al.

Most existing algorithms for depth estimation from single monocular images need large quantities of metric groundtruth depths for supervised learning. We show that relative depth can be an informative cue for metric depth estimation and can be easily obtained from vast stereo videos. Acquiring metric depths from stereo videos is sometimes impracticable due to the absence of camera parameters. In this paper, we propose to improve the performance of metric depth estimation with relative depths collected from stereo movie videos using existing stereo matching algorithm. We introduce a new "Relative Depth in Stereo" (RDIS) dataset densely labelled with relative depths. We first pretrain a ResNet model on our RDIS dataset. Then we finetune the model on RGB-D datasets with metric ground-truth depths. During our finetuning, we formulate depth estimation as a classification task. This re-formulation scheme enables us to obtain the confidence of a depth prediction in the form of probability distribution. With this confidence, we propose an information gain loss to make use of the predictions that are close to ground-truth. We evaluate our approach on both indoor and outdoor benchmark RGB-D datasets and achieve state-of-the-art performance.