Search

Your search keyword '"Luo, Yihong"' showing total 42 results
Start Over Author "Luo, Yihong"
42 results on '"Luo, Yihong"'

1. You Only Sample Once: Taming One-Step Text-to-Image Synthesis by Self-Cooperative Diffusion GANs

Author

Luo, Yihong, Chen, Xiaolong, Qu, Xinghua, and Tang, Jing

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

We introduce YOSO, a novel generative model designed for rapid, scalable, and high-fidelity one-step image synthesis. YOSO integrates the diffusion process with GANs to achieve the best of two worlds. Specifically, we smooth the distribution by the denoising generator itself, performing self-cooperative learning. We show that our method can serve as a one-step generation model training from scratch with competitive performance. Moreover, we show that our method can be extended to finetune pre-trained text-to-image diffusion for high-quality one-step text-to-image synthesis even with LoRA fine-tuning. In particular, we provide the first diffusion transformer that can generate images in one step trained on 512 resolution, with the capability of adapting to 1024 resolution without extra explicit training. Our code is provided at https://github.com/Luo-Yihong/YOSO, Comment: Early version

Published
2024

2. Energy-Calibrated VAE with Test Time Free Lunch

Author

Luo, Yihong, Qiu, Siya, Tao, Xingjian, Cai, Yujun, and Tang, Jing

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

In this paper, we propose a novel generative model that utilizes a conditional Energy-Based Model (EBM) for enhancing Variational Autoencoder (VAE), termed Energy-Calibrated VAE (EC-VAE). Specifically, VAEs often suffer from blurry generated samples due to the lack of a tailored training on the samples generated in the generative direction. On the other hand, EBMs can generate high-quality samples but require expensive Markov Chain Monte Carlo (MCMC) sampling. To address these issues, we introduce a conditional EBM for calibrating the generative direction of VAE during training, without requiring it for the generation at test time. In particular, we train EC-VAE upon both the input data and the calibrated samples with adaptive weight to enhance efficacy while avoiding MCMC sampling at test time. Furthermore, we extend the calibration idea of EC-VAE to variational learning and normalizing flows, and apply EC-VAE to an additional application of zero-shot image restoration via neural transport prior and range-null theory. We evaluate the proposed method with two applications, including image generation and zero-shot image restoration, and the experimental results show that our method achieves competitive performance over single-step non-adversarial generation. Our code is available at https://github.com/DJ-LYH/EC-VAE., Comment: ECCV 2024. Code is available at https://github.com/DJ-LYH/EC-VAE

Published
2023

3. LSGNN: Towards General Graph Neural Network in Node Classification by Local Similarity

Author

Chen, Yuhan, Luo, Yihong, Tang, Jing, Yang, Liang, Qiu, Siya, Wang, Chuan, and Cao, Xiaochun

Subjects
Computer Science - Machine Learning, Computer Science - Artificial Intelligence
Abstract

Heterophily has been considered as an issue that hurts the performance of Graph Neural Networks (GNNs). To address this issue, some existing work uses a graph-level weighted fusion of the information of multi-hop neighbors to include more nodes with homophily. However, the heterophily might differ among nodes, which requires to consider the local topology. Motivated by it, we propose to use the local similarity (LocalSim) to learn node-level weighted fusion, which can also serve as a plug-and-play module. For better fusion, we propose a novel and efficient Initial Residual Difference Connection (IRDC) to extract more informative multi-hop information. Moreover, we provide theoretical analysis on the effectiveness of LocalSim representing node homophily on synthetic graphs. Extensive evaluations over real benchmark datasets show that our proposed method, namely Local Similarity Graph Neural Network (LSGNN), can offer comparable or superior state-of-the-art performance on both homophilic and heterophilic graphs. Meanwhile, the plug-and-play model can significantly boost the performance of existing GNNs. Our code is provided at https://github.com/draym28/LSGNN., Comment: The first two authors contributed equally to this work; IJCAI23

Published
2023

6. TO-FLOW: Efficient Continuous Normalizing Flows with Temporal Optimization adjoint with Moving Speed

Author

Du, Shian, Luo, Yihong, Chen, Wei, Xu, Jian, and Zeng, Delu

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Continuous normalizing flows (CNFs) construct invertible mappings between an arbitrary complex distribution and an isotropic Gaussian distribution using Neural Ordinary Differential Equations (neural ODEs). It has not been tractable on large datasets due to the incremental complexity of the neural ODE training. Optimal Transport theory has been applied to regularize the dynamics of the ODE to speed up training in recent works. In this paper, a temporal optimization is proposed by optimizing the evolutionary time for forward propagation of the neural ODE training. In this appoach, we optimize the network weights of the CNF alternately with evolutionary time by coordinate descent. Further with temporal regularization, stability of the evolution is ensured. This approach can be used in conjunction with the original regularization approach. We have experimentally demonstrated that the proposed approach can significantly accelerate training without sacrifying performance over baseline models., Comment: To be published in CVPR 2022

Published
2022

11. Ecological Administration Contract Design for Enterprise and Local Government Based on Bottom-Line Thinking

Author

Lin, Hongwei, Li, Chenggang, Luo, Yihong, Akan, Ozgur, Editorial Board Member, Bellavista, Paolo, Editorial Board Member, Cao, Jiannong, Editorial Board Member, Coulson, Geoffrey, Editorial Board Member, Dressler, Falko, Editorial Board Member, Ferrari, Domenico, Editorial Board Member, Gerla, Mario, Editorial Board Member, Kobayashi, Hisashi, Editorial Board Member, Palazzo, Sergio, Editorial Board Member, Sahni, Sartaj, Editorial Board Member, Shen, Xuemin (Sherman), Editorial Board Member, Stan, Mircea, Editorial Board Member, Jia, Xiaohua, Editorial Board Member, Zomaya, Albert Y., Editorial Board Member, Song, Houbing, editor, and Jiang, Dingde, editor

Published
2021
Full Text
View/download PDF

14. Peripheral Soluble Immune Checkpoint-Related Proteins Were Associated with Survival and Treatment Efficacy of Osteosarcoma Patients, a Cohort Study.

Author

Li, Binghao, Wang, Qinchuan, Luo, Yihong, Wang, Sicong, Pan, Sai, Zhao, Wenting, Ye, Zhaoming, and Wu, Xifeng

Subjects
OSTEOSARCOMA, MEDICAL logic, PEARSON correlation (Statistics), PREDICTION models, T-test (Statistics), STATISTICAL significance, RESEARCH funding, IMMUNOTHERAPY, MULTIPLE regression analysis, CANCER patients, TREATMENT effectiveness, MULTIVARIATE analysis, MANN Whitney U Test, CHI-squared test, DESCRIPTIVE statistics, METASTASIS, LOG-rank test, KAPLAN-Meier estimator, LUNG tumors, STATISTICS, PROGRESSION-free survival, DATA analysis software, IMMUNE checkpoint proteins, OVERALL survival, BIOMARKERS, PROPORTIONAL hazards models, DISEASE progression
Abstract

Simple Summary: Osteosarcoma is one of the most lethal bone tumors worldwide. Immune checkpoint blockades have achieved significant success in solid tumors; however, their role in osteosarcoma remains obscure. Therefore, we aim to explore the clinical significance of soluble immune checkpoint-related proteins in osteosarcoma in this study. We identified four soluble immune checkpoint-related proteins as predictors of progress-free survival and lung metastasis-free survival of osteosarcoma patients. Our findings indicated that soluble immune checkpoint-related proteins could be promising biomarkers for the outcomes and immunotherapy of osteosarcoma. Background: The immune checkpoint blockade remains obscure in osteosarcoma (OS). We aim to explore the clinical significance of soluble immune checkpoint (ICK)-related proteins in OS. Methods: We profiled 14 soluble ICK-related proteins (BTLA, GITR, HVEM, IDO, LAG-3, PD-1, PD-L1, PD-L2, TIM-3, CD28, CD80, CD137, CD27, and CTLA-4) in the plasma of 76 OS patients and matched controls. We evaluated the associations between the biomarkers and the risk of OS using unconditional multivariate logistic regression. The multivariate Cox model was utilized to develop the prediction model of OS. Immune subtypes were established from the identified biomarkers. Transcriptional data from GEO were analyzed to elucidate potential mechanisms. Results: We found that sTIM3, sCD137, sIDO, and sCTLA4 were significantly correlated with OS risk (all p < 0.05). sBTLA, sPDL2, and sCD27 were significantly associated with the risk of lung metastasis, whereas sBTLA and sTIM3 were associated with the risk of disease progression. We also established an immune subtype based on sBTLA, sPD1, sTIM3, and sPDL2. Patients in the sICK-type2 subtype had significantly decreased progression-free survival (PFS) and lung metastasis-free survival (LMFS) than those in the sICK-type1 subtype (log-rank p = 2.8 × 10−2, 1.7 × 10−2, respectively). Interestingly, we found that the trend of LMFS and PFS in the subtypes of corresponding ICK genes' expression was opposite to the results in the blood (log-rank p = 2.6 × 10−4, 9.5 × 10−4, respectively). Conclusion: Four soluble ICK-related proteins were associated with the survival of OS patients. Soluble ICK-related proteins could be promising biomarkers for the outcomes and immunotherapy of OS patients, though more research is warranted. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

21. Eland: An Efficient Lightweight Anonymous Authentication Protocol Applied to Digital Rights Management System

Author

Qing Fan Qing Fan, Jianhua Chen Qing Fan, Yihong Wen Jianhua Chen, and Min Luo Yihong Wen

Subjects
Computer Networks and Communications, ComputingMilieux_LEGALASPECTSOFCOMPUTING, Software
Abstract

A digital rights management (DRM) system provides the function of packaging raw content into an appropriate distribution form, protection for content temper-proof transmission and circumvention of unauthorized use. The digital rights contents in DRM system improve people’s spiritual quality and create huge market value. However, in the sharing of digital contents, malicious adversaries could bring security and privacy issues even cause infringement of copyright. To solve these problems, valid user identity authentication and secret key distribution to legal users is an efficient method. Nevertheless, the existing authentication schemes for DRM system have problems of security weakness or efficiency to be enhanced. Thereby, we cryptanalyze these protocols’ vulnerabilities and propose a new lightweight anonymous authentication protocol called Eland that can be efficiently applied to DRM system. We compare our protocol with six DRM-related authentication protocols from the aspects of security properties, computation expense and communication cost. Comparisons results indicate that our protocol achieves a well trade-off between security and performance. Specifically, our protocol reduces 37.5% communication cost and satisfies more security requirements than Yu et al.’s scheme. &nbsp

Published
2022

34. Second Harmonic Weighted Reconstruction for Non-Contact Monitoring Heart Rate.

Author

Zheng, Pengwei, Zheng, Chundi, Li, Xiaoxia, Chen, Huihui, Wang, Aiguo, and Luo, Yihong

Abstract

Achieving high-accuracy and non-contact heart rate monitoring via radar is a non-trivial task under the interference from respiration and its harmonics. In order to eliminate the interference, we propose a weighted reconstruction method of the second harmonic signal of the heartbeat for the heart rate estimation. The proposed method combines the Variational Modal Decomposition (VMD) algorithm with the prior information about the second harmonic of the heartbeat. Namely, after a sophisticated endeavor for preprocessing, the chest wall displacement signal is decomposed into several quasi-orthogonal components by VMD and then those components whose spectrum are within the second harmonic frequency band of the heartbeat are selected to reconstruct the second harmonic signal of the heartbeat. For highlighting more likely second harmonic components and suppressing undesired noise components, the energy of each selected component is used as a weight to adjust its proportion in the reconstructed signal. Consequently, the heart rate estimation is achieved indirectly by estimating the spectral peak of the second harmonic signal. The proposed method not only avoids the strong interference from the respiratory signal and its higher harmonics on heart rate estimation, but also predigests the parameter setting in the spectral peak estimation model. Experimental results from 15 subjects demonstrate that the proposed method can achieve higher measurement accuracy compared with other several methods based on a 77 GHz FMCW radar. [ABSTRACT FROM AUTHOR]

Published
2022
Full Text
View/download PDF

37. Study of Improved Algorithms Based on TRIZ Ideality Level

Author

Luo Yihong, Chen Ting, and Shao Yunfei

Subjects
Range (mathematics), General Computer Science, law, General Mathematics, TRIZ, Innovation system, Inventive problem solving, Algorithm, Expression (mathematics), Interpretation (model theory), Mathematics, law.invention
Abstract

The traditional TRIZ(Teoriya Resheniya Izobretatelskikh Zadatch, Theory of Inventive Problem Solving) ideality level algorithm is improved, on the basis of interpretation of the meaning of ideality level, the concept of ideality level is redefined, the expression, range, domain, and application range of ideality level are explored; by adding weight parameters in the ideality level calculation formula, the relative importance of specific functions is reflected, the concept of idealized innovation is proposed, the correlations among each innovation stage in the total innovation system are established, and the specific application of ideality level is illustrated with the technological evolution process of washing machine as an example.

Published
2013

39. Long noncoding RNA ENST00000508435promotes migration of breast cancer via FXR 1

Author

Li, Luying, Jin, Youping, Wang, Xue, Wang, Li, Sun, Yangbai, Luo, Yihong, Ni, Xiaojian, Lu, Qi, Wei, Wenbo, Zhi, Xiuling, Yu, Jerry, Zhu, Wei, and Zhou, Ping

Abstract

ABSTRACTLncRNA plays a critical role in tumor progression. However, the role it executes in breast cancer is still unclear. Here, we report a newly discovered lncRNA, ENST00000508435, which could be remarkably up-regulated in breast cancer cells and tissues. We found that the expression of ENST00000508435was positively correlated with tumor size, lymph node metastasis and HER2. More interesting, overexpression of ENST00000508435significantly increased cell migration, while specific knockdown led to the opposite. RNA pull-down and RNA immunoprecipitation assays demonstrated that ENST00000508435could directly bind to FXR1 to promote tumor metastasis. ENST00000508435and FXR1 were positively correlated. FXR1 was also significantly up-regulated in breast tumors. Taken together, we propose that ENST00000508435regulates FXR1 to promote breast cancer metastasis.

Published
2021
Full Text
View/download PDF

40. Dynamics of triple helix relations in the development of cleaner technologies: case of a Chinese power equipment manufacturer.

Author

Yang, Yan, Etzkowitz, Henry, Yin, Shoujun, and Luo, Yihong

Subjects
ECONOMIC development, SUSTAINABLE development, ENERGY consumption, RENEWABLE energy sources, ELECTRIC power production
Abstract

This paper analyses the 60-year-long dynamics of triple helix relations in the development of cleaner power generation technologies of a large state-owned Chinese power equipment manufacturer. It opens the black box of the firm and provides a whole picture of the firm's triple helix relations and the changes in the process. It addresses two research questions: How did the triple helix relations of the case firm evolve over the different stages of technology development? What's the special role of government in the evolution of triple helix relations in China? A heterodox top-down 'coordinated' model of triple helix relations is identified. [ABSTRACT FROM AUTHOR]

Published
2019
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results