arXiv subjects

Yu Liu

Publications and source records attributed to Yu Liu.

At least 613 records · Page 34Linked to original sources

arXiv2024

Rethinking the Spatial Inconsistency in Classifier-Free Diffusion Guidance

Classifier-Free Guidance (CFG) has been widely used in text-to-image diffusion models, where the CFG scale is introduced to control the strength of text guidance on the whole image space. However, we argue that a global CFG scale results in spatial inconsistency on varying semantic strengths and suboptimal image quality. To address this problem, we present a novel approach, Semantic-aware Classifier-Free Guidance (S-CFG), to customize the guidance degrees for different semantic units in text-to-image diffusion models. Specifically, we first design a training-free semantic segmentation method to partition the latent image into relatively independent semantic regions at each denoising step. In particular, the cross-attention map in the denoising U-net backbone is renormalized for assigning each patch to the corresponding token, while the self-attention map is used to complete the semantic regions. Then, to balance the amplification of diverse semantic units, we adaptively adjust the CFG scales across different semantic regions to rescale the text guidance degrees into a uniform level. Finally, extensive experiments demonstrate the superiority of S-CFG over the original CFG strategy on various text-to-image diffusion models, without requiring any extra training cost. our codes are available at https://github.com/SmilesDZgk/S-CFG.

cs.CV↗

arXiv2024

Deep-Learning-Based Channel Estimation for IRS-Assisted ISAC System

Integrated sensing and communication (ISAC) and intelligent reflecting surface (IRS) are viewed as promising technologies for future generations of wireless networks. This paper investigates the channel estimation problem in an IRS-assisted ISAC system. A deep-learning framework is proposed to estimate the sensing and communication (S&C) channels in such a system. Considering different propagation environments of the S&C channels, two deep neural network (DNN) architectures are designed to realize this framework. The first DNN is devised at the ISAC base station for estimating the sensing channel, while the second DNN architecture is assigned to each downlink user equipment to estimate its communication channel. Moreover, the input-output pairs to train the DNNs are carefully designed. Simulation results show the superiority of the proposed estimation approach compared to the benchmark scheme under various signal-to-noise ratio conditions and system parameters.

eess.SP↗

arXiv2024

Extreme Learning Machine-based Channel Estimation in IRS-Assisted Multi-User ISAC System

Multi-user integrated sensing and communication (ISAC) assisted by intelligent reflecting surface (IRS) has been recently investigated to provide a high spectral and energy efficiency transmission. This paper proposes a practical channel estimation approach for the first time to an IRS-assisted multiuser ISAC system. The estimation problem in such a system is challenging since the sensing and communication (SAC) signals interfere with each other, and the passive IRS lacks signal processing ability. A two-stage approach is proposed to transfer the overall estimation problem into sub-ones, successively including the direct and reflected channels estimation. Based on this scheme, the ISAC base station (BS) estimates all the SAC channels associated with the target and uplink users, while each downlink user estimates the downlink communication channels individually. Considering a low-cost demand of the ISAC BS and downlink users, the proposed two-stage approach is realized by an efficient neural network (NN) framework that contains two different extreme learning machine (ELM) structures to estimate the above SAC channels. Moreover, two types of input-output pairs to train the ELMs are carefully devised, which impact the estimation accuracy and computational complexity under different system parameters. Simulation results reveal a substantial performance improvement achieved by the proposed ELM-based approach over the least-squares and NN-based benchmarks, with reduced training complexity and faster training speed.

eess.SP↗

arXiv2024

Deep-Learning Channel Estimation for IRS-Assisted Integrated Sensing and Communication System

Integrated sensing and communication (ISAC), and intelligent reflecting surface (IRS) are envisioned as revolutionary technologies to enhance spectral and energy efficiencies for next wireless system generations. For the first time, this paper focuses on the channel estimation problem in an IRS-assisted ISAC system. This problem is challenging due to the lack of signal processing capacity in passive IRS, as well as the presence of mutual interference between sensing and communication (SAC) signals in ISAC systems. A three-stage approach is proposed to decouple the estimation problem into sub-ones, including the estimation of the direct SAC channels in the first stage, reflected communication channel in the second stage, and reflected sensing channel in the third stage. The proposed three-stage approach is based on a deep-learning framework, which involves two different convolutional neural network (CNN) architectures to estimate the channels at the full-duplex ISAC base station. Furthermore, two types of input-output pairs to train the CNNs are carefully designed, which affect the estimation performance under various signal-to-noise ratio conditions and system parameters. Simulation results validate the superiority of the proposed estimation approach compared to the least-squares baseline scheme, and its computational complexity is also analyzed.

eess.SP↗

arXiv2024

Genetic Quantization-Aware Approximation for Non-Linear Operations in Transformers

Non-linear functions are prevalent in Transformers and their lightweight variants, incurring substantial and frequently underestimated hardware costs. Previous state-of-the-art works optimize these operations by piece-wise linear approximation and store the parameters in look-up tables (LUT), but most of them require unfriendly high-precision arithmetics such as FP/INT 32 and lack consideration of integer-only INT quantization. This paper proposed a genetic LUT-Approximation algorithm namely GQA-LUT that can automatically determine the parameters with quantization awareness. The results demonstrate that GQA-LUT achieves negligible degradation on the challenging semantic segmentation task for both vanilla and linear Transformer models. Besides, proposed GQA-LUT enables the employment of INT8-based LUT-Approximation that achieves an area savings of 81.3~81.7% and a power reduction of 79.3~80.2% compared to the high-precision FP/INT 32 alternatives. Code is available at https:// github.com/PingchengDong/GQA-LUT.

cs.LG↗

arXiv2024

Phase evolution of Ce-based heavy-fermion superconductors under compression: a combined first-principle and effective-model study

In many Ce-based superconductors, superconducting (SC) phases emerge or can be tuned in proximity to the antiferromagnetic (AF) quantum critical point (QCP), but so far the explicit phase evolution near the QCP lack theoretical understanding. Here, by combing the density functional theory plus dynamical mean-field theory (DFT+DMFT) with effective-model calculations, we provide a theoretical description for Ce-based superconductors under compression. DFT+DMFT calculations for the normal states reveal that the Kondo hybridizations are significantly enhanced under compression, while the initially localized $f$ electrons become fully itinerant via localized-itinerant crossover. We then construct an effective model and show that with the extracted Kondo coupling and RKKY exchange strengths from first-principle calculations, the ground-state phases of these materials can be properly predicted. We also show that the coexistence of magnetic correlation and Kondo hybridization can drive AF+SC coexisting state in narrow compression region. Under compression, competition between Kondo and RKKY interactions can drive successive transitions, from AF phase to AF+SC coexisting phase, then to paramagnetic SC phase via an AF transition which generates the QCP, and finally to normal Kondo paramagnetic (KP) phase through an SC-KP transition induced by the localized-itinerant crossover. Our study gives proper explanation to the pressure-induced QCP and SC-KP transition, and to the phase evolution in pressured Ce-based superconductors, and can help to understand the SC states around the ferromagnetic quantum transition points in uranium-based superconductors.

cond-mat.str-el↗

arXiv2024

Weakly Supervised AUC Optimization: A Unified Partial AUC Approach

Since acquiring perfect supervision is usually difficult, real-world machine learning tasks often confront inaccurate, incomplete, or inexact supervision, collectively referred to as weak supervision. In this work, we present WSAUC, a unified framework for weakly supervised AUC optimization problems, which covers noisy label learning, positive-unlabeled learning, multi-instance learning, and semi-supervised learning scenarios. Within the WSAUC framework, we first frame the AUC optimization problems in various weakly supervised scenarios as a common formulation of minimizing the AUC risk on contaminated sets, and demonstrate that the empirical risk minimization problems are consistent with the true AUC. Then, we introduce a new type of partial AUC, specifically, the reversed partial AUC (rpAUC), which serves as a robust training objective for AUC maximization in the presence of contaminated labels. WSAUC offers a universal solution for AUC optimization in various weakly supervised scenarios by maximizing the empirical rpAUC. Theoretical and experimental results under multiple settings support the effectiveness of WSAUC on a range of weakly supervised AUC optimization tasks.

cs.LG↗

arXiv2024

TagAlign: Improving Vision-Language Alignment with Multi-Tag Classification

The crux of learning vision-language models is to extract semantically aligned information from visual and linguistic data. Existing attempts usually face the problem of coarse alignment, e.g., the vision encoder struggles in localizing an attribute-specified object. In this work, we propose an embarrassingly simple approach to better align image and text features with no need of additional data formats other than image-text pairs. Concretely, given an image and its paired text, we manage to parse objects (e.g., cat) and attributes (e.g., black) from the description, which are highly likely to exist in the image. It is noteworthy that the parsing pipeline is fully automatic and thus enjoys good scalability. With these parsed semantics as supervision signals, we can complement the commonly used image-text contrastive loss with the multi-tag classification loss. Extensive experimental results on a broad suite of semantic segmentation datasets substantiate the average 5.2\% improvement of our framework over existing alternatives. Furthermore, the visualization results indicate that attribute supervision makes vision-language models accurately localize attribute-specified objects. Project page can be found at https://qinying-liu.github.io/Tag-Align.

cs.CV↗

arXiv2024

Check, Locate, Rectify: A Training-Free Layout Calibration System for Text-to-Image Generation

Diffusion models have recently achieved remarkable progress in generating realistic images. However, challenges remain in accurately understanding and synthesizing the layout requirements in the textual prompts. To align the generated image with layout instructions, we present a training-free layout calibration system SimM that intervenes in the generative process on the fly during inference time. Specifically, following a "check-locate-rectify" pipeline, the system first analyses the prompt to generate the target layout and compares it with the intermediate outputs to automatically detect errors. Then, by moving the located activations and making intra- and inter-map adjustments, the rectification process can be performed with negligible computational overhead. To evaluate SimM over a range of layout requirements, we present a benchmark SimMBench that compensates for the lack of superlative spatial relations in existing datasets. And both quantitative and qualitative results demonstrate the effectiveness of the proposed SimM in calibrating the layout inconsistencies. Our project page is at https://simm-t2i.github.io/SimM.

cs.CV↗

arXiv2024

The effect of surface oxidation and crystal thickness on magnetic properties and magnetic domain structures of Cr2Ge2Te6

Van der Waals (vdW) magnetic materials such as Cr2Ge2Te6 (CGT) show promise for novel memory and logic applications. This is due to their broadly tunable magnetic properties and the presence of topological magnetic features such as skyrmionic bubbles. A systematic study of thickness and oxidation effects on magnetic domain structures is important for designing devices and vdW heterostructures for practical applications. Here, we investigate thickness effects on magnetic properties, magnetic domains, and bubbles in oxidation-controlled CGT crystals. We find that CGT exposed to ambient conditions for 5 days forms an oxide layer approximately 5 nm thick. This oxidation leads to a significant increase in the oxidation state of the Cr ions, indicating a change in local magnetic properties. This is supported by real space magnetic texture imaging through Lorentz transmission electron microscopy. By comparing the thickness dependent saturation field of oxidized and pristine crystals, we find that oxidation leads to a non-magnetic surface layer which is thicker than the oxide layer alone. We also find that the stripe domain width and skyrmionic bubble size are strongly affected by the crystal thickness in pristine crystals. These findings underscore the impact of thickness and surface oxidation on the properties of CGT such as saturation field and domain/skyrmionic bubble size and suggest a pathway for manipulating magnetic properties through a controlled oxidation process.

cond-mat.mtrl-sci↗

arXiv2024

Quantum spin driven Yu-Shiba-Rusinov multiplets and fermion-parity-preserving phase transition in K$_3$C$_{60}$

Magnetic impurities in superconductors are of increasing interest due to emergent Yu-Shiba-Rusinov (YSR) states and Majorana zero modes for fault-tolerant quantum computation. However, a direct relationship between the YSR multiple states and magnetic anisotropy splitting of quantum impurity spins remains poorly characterized. By using scanning tunneling microscopy, we resolve systematically individual transition-metal (Fe, Cr and Ni) impurities induced YSR multiplets as well as their Zeeman effects in K$_3$C$_{60}$ superconductor. The YSR multiplets show identical $d$ orbital-like wave functions that are symmetry-mismatched to the threefold K$_3$C$_{60}$(111) host surface, breaking point-group symmetries of the spatial distribution of YSR bound states in real space. Remarkably, we identify an unprecedented fermion-parity-preserving quantum phase transition between ground states with opposite signs of the uniaxial magnetic anisotropy that can be manipulated by an external magnetic field. These findings can be readily understood in terms of anisotropy splitting of quantum impurity spins, and thus elucidate the intricate interplay between the magnetic anisotropy and YSR multiplets.

cond-mat.supr-con↗

arXiv2024

Be-Your-Outpainter: Mastering Video Outpainting through Input-Specific Adaptation

Video outpainting is a challenging task, aiming at generating video content outside the viewport of the input video while maintaining inter-frame and intra-frame consistency. Existing methods fall short in either generation quality or flexibility. We introduce MOTIA Mastering Video Outpainting Through Input-Specific Adaptation, a diffusion-based pipeline that leverages both the intrinsic data-specific patterns of the source video and the image/video generative prior for effective outpainting. MOTIA comprises two main phases: input-specific adaptation and pattern-aware outpainting. The input-specific adaptation phase involves conducting efficient and effective pseudo outpainting learning on the single-shot source video. This process encourages the model to identify and learn patterns within the source video, as well as bridging the gap between standard generative processes and outpainting. The subsequent phase, pattern-aware outpainting, is dedicated to the generalization of these learned patterns to generate outpainting outcomes. Additional strategies including spatial-aware insertion and noise travel are proposed to better leverage the diffusion model's generative prior and the acquired video patterns from source videos. Extensive evaluations underscore MOTIA's superiority, outperforming existing state-of-the-art methods in widely recognized benchmarks. Notably, these advancements are achieved without necessitating extensive, task-specific tuning.

cs.CV↗

arXiv2024

Mobility Census for monitoring rapid urban development

Monitoring urban structure and development requires high-quality data at high spatiotemporal resolution. While traditional censuses have provided foundational insights into demographic and socioeconomic aspects of urban life, their pace may not always align with the pace of urban development. To complement these traditional methods, we explore the potential of analyzing alternative big-data sources, such as human mobility data. However, these often noisy and unstructured big data pose new challenges. Here we propose a method to extract meaningful explanatory variables and classifications from such data. Using movement data from Beijing, which are produced as a byproduct of mobile communication, we show that meaningful features can be extracted, revealing, for example, the emergence and absorption of subcentres. This method allows the analysis of urban dynamics at a high spatial resolution (here, 500m) and near real-time frequency, and high computational efficiency, which is especially suitable for tracing event-driven mobility changes and their impact on urban structures.

physics.soc-ph↗

arXiv2024

SmartRefine: A Scenario-Adaptive Refinement Framework for Efficient Motion Prediction

Predicting the future motion of surrounding agents is essential for autonomous vehicles (AVs) to operate safely in dynamic, human-robot-mixed environments. Context information, such as road maps and surrounding agents' states, provides crucial geometric and semantic information for motion behavior prediction. To this end, recent works explore two-stage prediction frameworks where coarse trajectories are first proposed, and then used to select critical context information for trajectory refinement. However, they either incur a large amount of computation or bring limited improvement, if not both. In this paper, we introduce a novel scenario-adaptive refinement strategy, named SmartRefine, to refine prediction with minimal additional computation. Specifically, SmartRefine can comprehensively adapt refinement configurations based on each scenario's properties, and smartly chooses the number of refinement iterations by introducing a quality score to measure the prediction quality and remaining refinement potential of each scenario. SmartRefine is designed as a generic and flexible approach that can be seamlessly integrated into most state-of-the-art motion prediction models. Experiments on Argoverse (1 & 2) show that our method consistently improves the prediction accuracy of multiple state-of-the-art prediction models. Specifically, by adding SmartRefine to QCNet, we outperform all published ensemble-free works on the Argoverse 2 leaderboard (single agent track) at submission. Comprehensive studies are also conducted to ablate design choices and explore the mechanism behind multi-iteration refinement. Codes are available at https://github.com/opendilab/SmartRefine/

cs.CV↗

arXiv2024

FouriScale: A Frequency Perspective on Training-Free High-Resolution Image Synthesis

In this study, we delve into the generation of high-resolution images from pre-trained diffusion models, addressing persistent challenges, such as repetitive patterns and structural distortions, that emerge when models are applied beyond their trained resolutions. To address this issue, we introduce an innovative, training-free approach FouriScale from the perspective of frequency domain analysis. We replace the original convolutional layers in pre-trained diffusion models by incorporating a dilation technique along with a low-pass operation, intending to achieve structural consistency and scale consistency across resolutions, respectively. Further enhanced by a padding-then-crop strategy, our method can flexibly handle text-to-image generation of various aspect ratios. By using the FouriScale as guidance, our method successfully balances the structural integrity and fidelity of generated images, achieving an astonishing capacity of arbitrary-size, high-resolution, and high-quality generation. With its simplicity and compatibility, our method can provide valuable insights for future explorations into the synthesis of ultra-high-resolution images. The code will be released at https://github.com/LeonHLJ/FouriScale.

cs.CV↗

arXiv2024

Depth-induced Saliency Comparison Network for Diagnosis of Alzheimer's Disease via Jointly Analysis of Visual Stimuli and Eye Movements

Early diagnosis of Alzheimer's Disease (AD) is very important for following medical treatments, and eye movements under special visual stimuli may serve as a potential non-invasive biomarker for detecting cognitive abnormalities of AD patients. In this paper, we propose an Depth-induced saliency comparison network (DISCN) for eye movement analysis, which may be used for diagnosis the Alzheimers disease. In DISCN, a salient attention module fuses normal eye movements with RGB and depth maps of visual stimuli using hierarchical salient attention (SAA) to evaluate comprehensive saliency maps, which contain information from both visual stimuli and normal eye movement behaviors. In addition, we introduce serial attention module (SEA) to emphasis the most abnormal eye movement behaviors to reduce personal bias for a more robust result. According to our experiments, the DISCN achieves consistent validity in classifying the eye movements between the AD patients and normal controls.

cs.CV↗

arXiv2024

CSCNET: Class-Specified Cascaded Network for Compositional Zero-Shot Learning

Attribute and object (A-O) disentanglement is a fundamental and critical problem for Compositional Zero-shot Learning (CZSL), whose aim is to recognize novel A-O compositions based on foregone knowledge. Existing methods based on disentangled representation learning lose sight of the contextual dependency between the A-O primitive pairs. Inspired by this, we propose a novel A-O disentangled framework for CZSL, namely Class-specified Cascaded Network (CSCNet). The key insight is to firstly classify one primitive and then specifies the predicted class as a priori for guiding another primitive recognition in a cascaded fashion. To this end, CSCNet constructs Attribute-to-Object and Object-to-Attribute cascaded branches, in addition to a composition branch modeling the two primitives as a whole. Notably, we devise a parametric classifier (ParamCls) to improve the matching between visual and semantic embeddings. By improving the A-O disentanglement, our framework achieves superior results than previous competitive methods.

cs.CV↗

arXiv2024

RAPHAEL: Text-to-Image Generation via Large Mixture of Diffusion Paths

Text-to-image generation has recently witnessed remarkable achievements. We introduce a text-conditional image diffusion model, termed RAPHAEL, to generate highly artistic images, which accurately portray the text prompts, encompassing multiple nouns, adjectives, and verbs. This is achieved by stacking tens of mixture-of-experts (MoEs) layers, i.e., space-MoE and time-MoE layers, enabling billions of diffusion paths (routes) from the network input to the output. Each path intuitively functions as a "painter" for depicting a particular textual concept onto a specified image region at a diffusion timestep. Comprehensive experiments reveal that RAPHAEL outperforms recent cutting-edge models, such as Stable Diffusion, ERNIE-ViLG 2.0, DeepFloyd, and DALL-E 2, in terms of both image quality and aesthetic appeal. Firstly, RAPHAEL exhibits superior performance in switching images across diverse styles, such as Japanese comics, realism, cyberpunk, and ink illustration. Secondly, a single model with three billion parameters, trained on 1,000 A100 GPUs for two months, achieves a state-of-the-art zero-shot FID score of 6.61 on the COCO dataset. Furthermore, RAPHAEL significantly surpasses its counterparts in human evaluation on the ViLG-300 benchmark. We believe that RAPHAEL holds the potential to propel the frontiers of image generation research in both academia and industry, paving the way for future breakthroughs in this rapidly evolving field. More details can be found on a webpage: https://raphael-painter.github.io/.

cs.CV↗