المؤلفون: Chen, Lei, Meng, Yuan, Tang, Chen, Ma, Xinzhu, Jiang, Jingyan, Wang, Xin, Wang, Zhi, Zhu, Wenwu

مصطلحات موضوعية: Computer Science - Computer Vision and Pattern Recognition, Computer Science - Artificial Intelligence

الوصف: Recent advancements in diffusion models, particularly the trend of architectural transformation from UNet-based Diffusion to Diffusion Transformer (DiT), have significantly improved the quality and scalability of image synthesis. Despite the incredible generative quality, the large computational requirements of these large-scale models significantly hinder the deployments in real-world scenarios. Post-training Quantization (PTQ) offers a promising solution by compressing model sizes and speeding up inference for the pretrained models while eliminating model retraining. However, we have observed the existing PTQ frameworks exclusively designed for both ViT and conventional Diffusion models fall into biased quantization and result in remarkable performance degradation. In this paper, we find that the DiTs typically exhibit considerable variance in terms of both weight and activation, which easily runs out of the limited numerical representations. To address this issue, we devise Q-DiT, which seamlessly integrates three techniques: fine-grained quantization to manage substantial variance across input channels of weights and activations, an automatic search strategy to optimize the quantization granularity and mitigate redundancies, and dynamic activation quantization to capture the activation changes across timesteps. Extensive experiments on the ImageNet dataset demonstrate the effectiveness of the proposed Q-DiT. Specifically, when quantizing DiT-XL/2 to W8A8 on ImageNet 256x256, Q-DiT achieves a remarkable reduction in FID by 1.26 compared to the baseline. Under a W4A8 setting, it maintains high fidelity in image generation, showcasing only a marginal increase in FID and setting a new benchmark for efficient, high-quality quantization in diffusion transformers. Code is available at \href{https://github.com/Juanerx/Q-DiTTest}{https://github.com/Juanerx/Q-DiTTest}.

الوصول الحر: http://arxiv.org/abs/2406.17343Test

المؤلفون: Jiang, Qinting, Ye, Chuyang, Wei, Dongyan, Xue, Yuan, Jiang, Jingyan, Wang, Zhi

مصطلحات موضوعية: Computer Science - Machine Learning, Computer Science - Artificial Intelligence, Computer Science - Computer Vision and Pattern Recognition, Computer Science - Multimedia

الوصف: Despite progress, deep neural networks still suffer performance declines under distribution shifts between training and test domains, leading to a substantial decrease in Quality of Experience (QoE) for multimedia applications. Existing test-time adaptation (TTA) methods are challenged by dynamic, multiple test distributions within batches. This work provides a new perspective on analyzing batch normalization techniques through class-related and class-irrelevant features, our observations reveal combining source and test batch normalization statistics robustly characterizes target distributions. However, test statistics must have high similarity. We thus propose Discover Your Neighbours (DYN), the first backward-free approach specialized for dynamic TTA. The core innovation is identifying similar samples via instance normalization statistics and clustering into groups which provides consistent class-irrelevant representations. Specifically, Our DYN consists of layer-wise instance statistics clustering (LISC) and cluster-aware batch normalization (CABN). In LISC, we perform layer-wise clustering of approximate feature samples at each BN layer by calculating the cosine similarity of instance normalization statistics across the batch. CABN then aggregates SBN and TCN statistics to collaboratively characterize the target distribution, enabling more robust representations. Experimental results validate DYN's robustness and effectiveness, demonstrating maintained performance under dynamic data stream patterns.
Comment: 10 pages

الوصول الحر: http://arxiv.org/abs/2406.05413Test

المؤلفون: Sun, Haojun, Tang, Chen, Wang, Zhi, Meng, Yuan, jiang, Jingyan, Ma, Xinzhu, Zhu, Wenwu

مصطلحات موضوعية: Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning

الوصف: Diffusion models have emerged as preeminent contenders in the realm of generative models. Distinguished by their distinctive sequential generative processes, characterized by hundreds or even thousands of timesteps, diffusion models progressively reconstruct images from pure Gaussian noise, with each timestep necessitating full inference of the entire model. However, the substantial computational demands inherent to these models present challenges for deployment, quantization is thus widely used to lower the bit-width for reducing the storage and computing overheads. Current quantization methodologies primarily focus on model-side optimization, disregarding the temporal dimension, such as the length of the timestep sequence, thereby allowing redundant timesteps to continue consuming computational resources, leaving substantial scope for accelerating the generative process. In this paper, we introduce TMPQ-DM, which jointly optimizes timestep reduction and quantization to achieve a superior performance-efficiency trade-off, addressing both temporal and model optimization aspects. For timestep reduction, we devise a non-uniform grouping scheme tailored to the non-uniform nature of the denoising process, thereby mitigating the explosive combinations of timesteps. In terms of quantization, we adopt a fine-grained layer-wise approach to allocate varying bit-widths to different layers based on their respective contributions to the final generative performance, thus rectifying performance degradation observed in prior studies. To expedite the evaluation of fine-grained quantization, we further devise a super-network to serve as a precision solver by leveraging shared quantization results. These two design components are seamlessly integrated within our framework, enabling rapid joint exploration of the exponentially large decision space via a gradient-free evolutionary search algorithm.

الوصول الحر: http://arxiv.org/abs/2404.09532Test

المؤلفون: Jiang, Qinting, Xu, Xiaolong, Bilal, Muhammad, Crowcroft, Jon, Liu, Qi, Dou, Wanchun, Jiang, Jingyan

الوصف: Vehicular services aim to provide smart and timely services (e.g., collision warning) by taking the advantage of recent advances in artificial intelligence and employing task offloading techniques in mobile edge computing. In practice, the volume of vehicles in the Internet of Vehicles (IoV) often surges at a single location and renders the edge servers (ESs) severely overloaded, resulting in a very high delay in delivering the services. Therefore, it is of practical importance and urgency to coordinate the resources of ESs with bandwidth allocation for mitigating the occurrence of a spike traffic flow. For this challenge, existing work sought the periodicities of traffic flow by analyzing historical traffic data. However, the changes in traffic flow caused by sudden traffic conditions cannot be obtained from these periodicities. In this paper, we propose a distributed traffic flow forecasting and task offloading approach named TFFTO to optimize the execution time and power consumption in service processing. Specifically, graph attention networks (GATs) are leveraged to forecast future traffic flow in short-term and the traffic volume is utilized to estimate the number of services offloaded to the ESs in the subsequent period. With the estimate, the current load of the ESs is adjusted to ensure that the services can be handled in a timely manner. Potential game theory is adopted to determine the optimal service offloading strategy. Extensive experiments are conducted to evaluate our approach and the results validate our robust performance.

وصف الملف: text

العلاقة: https://eprints.lancs.ac.uk/id/eprint/216796/1Test/; https://eprints.lancs.ac.uk/id/eprint/216796/2Test/; https://eprints.lancs.ac.uk/id/eprint/216796/3Test/; https://eprints.lancs.ac.uk/id/eprint/216796/4Test/; https://eprints.lancs.ac.uk/id/eprint/216796/5Test/; https://eprints.lancs.ac.uk/id/eprint/216796/6/Potential_Game_based_Distributed_IoV_Service_Offloading_with_Graph_Attention_Networks_in_Mobile_Edge_Computing_1_.pdfTest; Jiang, Qinting and Xu, Xiaolong and Bilal, Muhammad and Crowcroft, Jon and Liu, Qi and Dou, Wanchun and Jiang, Jingyan (2024) Potential Game Based Distributed IoV Service Offloading With Graph Attention Networks in Mobile Edge Computing. IEEE Transactions on Intelligent Transportation Systems. ISSN 1524-9050

الإتاحة: https://eprints.lancs.ac.uk/id/eprint/216796Test/
https://eprints.lancs.ac.uk/id/eprint/216796/1Test/
https://eprints.lancs.ac.uk/id/eprint/216796/2Test/
https://eprints.lancs.ac.uk/id/eprint/216796/3Test/
https://eprints.lancs.ac.uk/id/eprint/216796/4Test/
https://eprints.lancs.ac.uk/id/eprint/216796/5Test/
https://eprints.lancs.ac.uk/id/eprint/216796/6/Potential_Game_based_Distributed_IoV_Service_Offloading_with_Graph_Attention_Networks_in_Mobile_Edge_Computing_1_.pdfTest

المؤلفون: Hu, Chenghao, Jiang, Jingyan, Wang, Zhi

مصطلحات موضوعية: Computer Science - Machine Learning, Computer Science - Distributed, Parallel, and Cluster Computing, Computer Science - Networking and Internet Architecture, Statistics - Machine Learning

الوصف: The emerging concern about data privacy and security has motivated the proposal of federated learning, which allows nodes to only synchronize the locally-trained models instead their own original data. Conventional federated learning architecture, inherited from the parameter server design, relies on highly centralized topologies and the assumption of large nodes-to-server bandwidths. However, in real-world federated learning scenarios the network capacities between nodes are highly uniformly distributed and smaller than that in a datacenter. It is of great challenges for conventional federated learning approaches to efficiently utilize network capacities between nodes. In this paper, we propose a model segment level decentralized federated learning to tackle this problem. In particular, we propose a segmented gossip approach, which not only makes full utilization of node-to-node bandwidth, but also has good training convergence. The experimental results show that even the training time can be highly reduced as compared to centralized federated learning.
Comment: Accepted to the 1st International Workshop on Federated Machine Learning for User Privacy and Data Confidentiality (FML'19)

الوصول الحر: http://arxiv.org/abs/1908.07782Test

المؤلفون: Li, Hongshan, Hu, Chenghao, Jiang, Jingyan, Wang, Zhi, Wen, Yonggang, Zhu, Wenwu

مصطلحات موضوعية: Computer Science - Distributed, Parallel, and Cluster Computing

الوصف: Recent years have witnessed a rapid growth of deep-network based services and applications. A practical and critical problem thus has emerged: how to effectively deploy the deep neural network models such that they can be executed efficiently. Conventional cloud-based approaches usually run the deep models in data center servers, causing large latency because a significant amount of data has to be transferred from the edge of network to the data center. In this paper, we propose JALAD, a joint accuracy- and latency-aware execution framework, which decouples a deep neural network so that a part of it will run at edge devices and the other part inside the conventional cloud, while only a minimum amount of data has to be transferred between them. Though the idea seems straightforward, we are facing challenges including i) how to find the best partition of a deep structure; ii) how to deploy the component at an edge device that only has limited computation power; and iii) how to minimize the overall execution latency. Our answers to these questions are a set of strategies in JALAD, including 1) A normalization based in-layer data compression strategy by jointly considering compression rate and model accuracy; 2) A latency-aware deep decoupling strategy to minimize the overall execution latency; and 3) An edge-cloud structure adaptation strategy that dynamically changes the decoupling for different network conditions. Experiments demonstrate that our solution can significantly reduce the execution latency: it speeds up the overall inference execution with a guaranteed model accuracy loss.
Comment: conference, copyright transfered to IEEE

الوصول الحر: http://arxiv.org/abs/1812.10027Test

المؤلفون: Song, Jian, Lu, Huimin, Liu, Meng, Hu, Hong, Jiang, Jingyan, Zhang, Lei, Li, Hui

المصدر: Small; 5/18/2024, Vol. 20 Issue 19, p1-10, 10p

المؤلفون: Huo, Xinwei, Shan, Xiaoyue, Pan, Qinghua, Cao, Zhanbo, Cong, Zichun, Song, Jian, Jiang, Jingyan, Ma, Qingfang, Jia, Linyan, Gao, Jianhua

المصدر: Sensors and Actuators B: Chemical ; volume 403, page 135184 ; ISSN 0925-4005

مصطلحات موضوعية: Materials Chemistry, Electrical and Electronic Engineering, Metals and Alloys, Surfaces, Coatings and Films, Condensed Matter Physics, Instrumentation, Electronic, Optical and Magnetic Materials

الإتاحة: https://doi.org/10.1016/j.snb.2023.135184Test
https://api.elsevier.com/content/article/PII:S0925400523019020?httpAccept=text/xmlTest
https://api.elsevier.com/content/article/PII:S0925400523019020?httpAccept=text/plainTest

المؤلفون: Qu, Wenjie, Li, Hua, Zhang, Teng, Zhang, Yuan, Ban, Yanli, Li, Ningfeng, Jiang, Jingyan, Xie, Juan, Shi, Wentian, Hao, Yiping, Li, Ruowen, Liu, Wei, Cui, Baoxia

المصدر: BMC Pregnancy and Childbirth ; volume 22, issue 1 ; ISSN 1471-2393

مصطلحات موضوعية: Obstetrics and Gynecology

الوصف: Background The aim of this study was to evaluate the effectiveness and safety of different treatment strategies for endogenic caesarean scar pregnancy (CSP) patients. Methods According to Vial’s standard, we defined endogenic-type CSP as (1) the gestational sac growing towards the uterine cavity and (2) a greater than 0.3 cm thickness of myometrial tissue at the caesarean scar. A total of 447 endogenic CSP patients out of 527 patients from 4 medical centres in China were enrolled in this study. A total of 120 patients were treated with methotrexate (MTX) followed by surgery, 106 received ultrasound-guided curettage directly and 221 received curettage combined with hysteroscopy. The clinical information and clinical outcomes of these patients were reviewed. Successful treatment was defined as (1) no additional treatment needed, (2) no retained mass of conception and (3) serum β subunit of human chorionic gonadotropin (β-hCG) level returning to a normal level within 4 weeks. The success rate was analysed based on these factors. Result Among 447 patients, no significant difference was observed in baseline characteristics between groups except for foetal heartbeat. The success rate was significantly different ( p <0.001) among the three groups. The highest success rate of 95.9% was noted in the hysteroscopy group, and the lowest success rate of 84.0% was noted in the curettage group. In addition, the MTX group reported the longest hospital stay and highest expenses, but the curettage group showed the shortest and lowest expenses, respectively. Nevertheless, no difference in blood loss was observed between the groups. Conclusion The combination of curettage and hysteroscopy represents the most effective strategy. Pretreatment with MTX did not result in better clinical outcomes. Ultrasound-guided curettage directly should not be considered a first-line treatment choice for endogenic CSP patients.

الإتاحة: https://doi.org/10.1186/s12884-022-04633-yTest

المؤلفون: Lu, Bo, Jiang, Jingyan, Li, Xiaoyu, Chen, Qingge, Qin, Jinling, Chen, Yun, Chen, Junping, Shen, Qing

المساهمون: Medical Scientific Research Foundation of Zhejiang Province, China

المصدر: BMJ Open ; volume 12, issue 5, page e057283 ; ISSN 2044-6055 2044-6055

الوصف: Objective We determined whether regional haemodynamics and perfusion index (PI) could be reliable indicators of a successful sciatic nerve block (SNB). Design Prospective observational trial. Setting A tertiary teaching hospital in China from April 2020 to August 2020. Participants We assessed 79 patients for eligibility to participate in this study. Nine patients were excluded for not meeting our inclusion criteria, and three patients were excluded due to missing measurements at all time points. Interventions The patients underwent SNB. Pulsed-wave Doppler and PI measurements were performed. Primary and secondary outcome measures The primary outcome measure was the diagnostic power of regional haemodynamic change and PI to predict successful SNB. The secondary outcome measure was the effect of SNB on the regional haemodynamics and PI in the lower extremity. Results We assessed 79 patients in this study and 67 patients available for the final analysis. The SNB was successful in 59 patients and failed in eight patients. There were no significant differences in demographic characteristics between the patients with successful and failed SNB. Starting from 10 min after SNB, the peak systolic velocity (PSV), end-diastolic velocity, time-averaged maximum velocity and time-averaged mean velocity of the anterior tibial artery and posterior tibial artery of patients in the successful SNB group were significantly higher than those in the failed SNB group (p<0.05). The PSV percentage increase at 10 min after SNB has great potential to predict the block success. The area under the receiver operating characteristic curve (AUC) values were 0.893 (95% CI 0.7809 to 1.000) and 0.880 (95% CI 0.7901 to 0.9699). The corresponding cut-off values were 19.22 and 35.88, respectively. The PI increased during 5–45 min intervals in patients with successful SNB. The AUC for the PI percentage increases at 10 min after SNB was 0.853 (95% CI 0.7035 to 1.000), with a cut-off value of 93.09. Conclusion The regional haemodynamic variables, ...

الإتاحة: https://doi.org/10.1136/bmjopen-2021-057283Test