A5075488752- Wireless Communication Networks Research
- Advanced Wireless Communication Techniques
- Advanced Wireless Network Optimization
- PAPR reduction in OFDM
- Image Processing Techniques and Applications
- Natural Language Processing Techniques
- Human Pose and Action Recognition
- Industrial Vision Systems and Defect Detection
- Satellite Communication Systems
- Advanced Neural Network Applications
- Multimodal Machine Learning Applications
- Advanced MIMO Systems Optimization
- Full-Duplex Wireless Communications
- Robotics and Sensor-Based Localization
- Advanced Vision and Imaging
- Domain Adaptation and Few-Shot Learning
- Topic Modeling
- Probabilistic and Robust Engineering Design
- IoT Networks and Protocols
- Robotic Path Planning Algorithms
- Cooperative Communication and Network Coding
- GNSS positioning and interference
- Anomaly Detection Techniques and Applications
- Coding theory and cryptography
- Inertial Sensor and Navigation
Beihang University
2021-2024
Xiangtan University
2021-2024
Sun Yat-sen University
2023
Beijing Institute of Technology
2021-2023
Shenzhen University
2023
University of Science and Technology Beijing
2022
Beijing University of Posts and Telecommunications
1999-2020
Nokia (Finland)
2017
University of Oulu
2002-2005
Knowledge Distillation (KD) for Convolutional Neural Network (CNN) is extensively studied as a way to boost the performance of small model. Recently, Vision Transformer (ViT) has achieved great success on many computer vision tasks and KD ViT also desired. However, besides output logit-based KD, other feature-based methods CNNs cannot be directly applied due huge structure gap. In this paper, we explore distillation ViT. Based nature feature maps in ViT, design series controlled experiments...
We present a strong object detector with encoder-decoder pretraining and finetuning. Our method, called Group DETR v2, is built upon vision transformer encoder ViT-Huge~\cite{dosovitskiy2020image}, variant DINO~\cite{zhang2022dino}, an efficient training method DETR~\cite{chen2022group}. The process consists of self-supervised finetuning ViT-Huge on ImageNet-1K, the Object365, finally it COCO. v2 achieves $\textbf{64.5}$ mAP COCO test-dev, establishes new SoTA leaderboard...
Image Signal Processor (ISP) is a crucial component in digital cameras that transforms sensor signals into images for us to perceive and understand. Existing ISP designs always adopt fixed architecture, e.g., several sequential modules connected rigid order. Such architecture may be suboptimal real-world applications, where camera sensors, scenes tasks are diverse. In this study, we propose novel Reconfigurable (ReconfigISP) whose parameters can automatically tailored specific data tasks....
Conditional masked language models (CMLM) have shown impressive progress in non-autoregressive machine translation (NAT). They learn the conditional model by predicting random subset target sentence. Based on CMLM framework, we introduce Multi-view Subset Regularization (MvSR), a novel regularization method to improve performance of NAT model. Specifically, MvSR consists two parts: (1) <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">shared...
Cooperative intelligent transport systems (ITS) and connected vehicles are foreseen to change the way mobility is conceived today. will lead improved road traffic safety efficiency also trigger innovation in infotainment area. These foster design of disruptive new business models for both telco automotive industries, triggering a profound impact society economy. However, before this can become reality, many technical challenges still need be solved. One important challenge relates provision...
In this paper, we present a novel simple genetic algorithm (GA) based multiuser detector for multicarrier code-division multiple-access (MC-CDMA) systems. More specifically, after getting the initial population from front-end frequency domain linear minimum mean-squared error (MMSE) equalizer, GA operators are applied to expand candidate list and maximum likelihood (ML) search is implemented in list. Compared MMSE receiver, proposed receiver can provide superior performance with very little...
The combination of code-division multiple-access (CDMA) and multicarrier modulation has been proposed for high data-rate wireless communication systems. In this paper, a novel practical technique based on alternative Gaussian approximation (AGA) determining the bit error rate (BER) maximal-ratio combining (MRC) receivers MC-CDMA systems in frequency selective Rayleigh fading channels is presented. impacts channel estimation errors are also taken into account both conventional MRC parallel...
Uncertainty exists widely in engineering design. As one of the key components design, uncertainty propagation and quantification has always been an important research topic. Polynomial chaos (PC) is a highly efficient method which studied applied. Therefore, this paper reviews recent advances PC method. First, fundamentals are introduced, including construction orthogonal polynomial basis calculation coefficients. Second, strategies such as truncation, sparse reconstruction, grid...
Spatio-temporal action detection methods locate human actions in both spatial and temporal dimension, which usually follow a two-stage structure. In this paper, We propose STD-TR, novel spatio-temporal framework with an end-to-end transformer STD-TR employs two branches to extract feature from video clip key frame concurrently, then sends the aggregated encoder-decoder. View as set matching prediction problem, learned object queries model relation of context, directly outputs all predictions...
We present a novel joint multiuser detection method based on sphere packing lattice decoding and semi-blind channel estimation for multicarrier code-division multiple-access (MC-CDMA) systems. After modelling MC-CDMA as lattice, low-complexity maximum-likelihood (ML) detection, algorithm, is applied to jointly detect all users. The impacts of errors are studied by incorporating subspace in the receiver. selection search radius complexity receiver also investigated. Another promising...
The Sign Language Production (SLP) project aims to automatically translate spoken languages into sign sequences. Our approach focuses on the transformation of gloss sequences their corresponding pose (G2P). In this paper, we present a novel solution for task by converting continuous space generation problem discrete sequence problem. We introduce Pose-VQVAE framework, which combines Variational Autoencoders (VAEs) with vector quantization produce latent representation Additionally, propose...
Multicarrier code division multiple access (MC-CDMA) is a promising technique that combines orthogonal frequency multiplexing (OFDM) with CDMA. In this paper, based on an alternative expression for the -function, characteristic function and Gaussian approximation, we present new practical determining bit error rate (BER) of multiuser MC-CDMA systems in frequency-selective Nakagami- fading channels. The results are applicable to employing coherent demodulation maximal ratio combining (MRC) or...
Abstract Satellite navigation positioning has become an indispensable component of everyday life, where precise pinpointing and rapid convergence are crucial in delivering timely accurate location information. However, due to the damping integer ambiguities system residual errors, Precise Point Positioning (PPP) implementation is a significant challenge. To address this, this paper proposes novel Carrier Phase Zero-Baseline Self-Differencing (CZS-PPP) technique its ionosphere-free fusion...
Arbitrary-resolution image generation still remains a challenging task in AIGC, as it requires handling varying resolutions and aspect ratios while maintaining high visual quality. Existing transformer-based diffusion methods suffer from quadratic computation cost limited resolution extrapolation capabilities, making them less effective for this task. In paper, we propose FlowDCN, purely convolution-based generative model with linear time memory complexity, that can efficiently generate...
The Sign Language Production (SLP) project aims to automatically translate spoken languages into sign sequences. Our approach focuses on the transformation of gloss sequences their corresponding pose (G2P). In this paper, we present a novel solution for task by converting continuous space generation problem discrete sequence problem. We introduce Pose-VQVAE framework, which combines Variational Autoencoders (VAEs) with vector quantization produce latent representation Additionally, propose...
Block encoding is a data input model commonly used in quantum computer. It technique that embeds matrix $A$ satisfying $\left\|A\right\| \leq 1$ into larger unitary $U_{A}$. We consider special structured matrices arising from generalized eigenvalue equations ocean acoustics. develop their block scheme and further improve it which results lower subnormalisations. And we discuss how to construct circuits of for the matrices. Two numerical examples are illustrate feasibility our schemes. The...
Multicarrier code division multiple access (MC-CDMA) has emerged as a promising air interface candidate for future wireless communication systems. With the help of characteristic function correlated Nakagami-m variables, performance an interleaved MC-CDMA with maximal-ratio combining (MRC) receiver and frequency selective fading channels is studied. Computer simulations demonstrate accuracy analysis. Based on analytical results, exhaustive search different subcarrier interleaving schemes,...
A novel multiuser detection scheme for space-time block coded multicarrier code-division multiple-access (STBC MC-CDMA) systems is proposed. More specifically, a semi-blind space-time-frequency minimum mean square error based parallel interference cancellation receiver (STF-MMSE/PIC) developed an STBC MC-CDMA system. The signal processing of this new detector jointly carried out in space, time and frequency domains, which leads to powerful technique combat the originating from different...
The following topics are dealt with: Internet of Things; satellite communication; resource allocation; telecommunication network routing; 5G mobile bandwidth optimisation; probability; software defined networking; traffic.