Search

Your search keyword '"Liu, Jiashuai"' showing total 32 results
Start Over Author "Liu, Jiashuai"
32 results on '"Liu, Jiashuai"'

1. Step-Audio: Unified Understanding and Generation in Intelligent Speech Interaction

Author

Huang, Ailin, Wu, Boyong, Wang, Bruce, Yan, Chao, Hu, Chen, Feng, Chengli, Tian, Fei, Shen, Feiyu, Li, Jingbei, Chen, Mingrui, Liu, Peng, Miao, Ruihang, You, Wang, Chen, Xi, Yang, Xuerui, Huang, Yechang, Zhang, Yuxiang, Gong, Zheng, Zhang, Zixin, Zhou, Hongyu, Sun, Jianjian, Li, Brian, Feng, Chengting, Wan, Changyi, Hu, Hanpeng, Wu, Jianchang, Zhen, Jiangjie, Ming, Ranchen, Yuan, Song, Zhang, Xuelin, Zhou, Yu, Li, Bingxin, Ma, Buyun, Wang, Hongyuan, An, Kang, Ji, Wei, Li, Wen, Wen, Xuan, Kong, Xiangwen, Ma, Yuankai, Liang, Yuanwei, Mou, Yun, Ahmidi, Bahtiyar, Wang, Bin, Li, Bo, Miao, Changxin, Xu, Chen, Wang, Chenrun, Shi, Dapeng, Sun, Deshan, Hu, Dingyuan, Sai, Dula, Liu, Enle, Huang, Guanzhe, Yan, Gulin, Wang, Heng, Jia, Haonan, Zhang, Haoyang, Gong, Jiahao, Guo, Junjing, Liu, Jiashuai, Liu, Jiahong, Feng, Jie, Wu, Jie, Wu, Jiaoren, Yang, Jie, Wang, Jinguo, Zhang, Jingyang, Lin, Junzhe, Li, Kaixiang, Xia, Lei, Zhou, Li, Zhao, Liang, Gu, Longlong, Chen, Mei, Wu, Menglin, Li, Ming, Li, Mingxiao, Li, Mingliang, Liang, Mingyao, Wang, Na, Hao, Nie, Wu, Qiling, Tan, Qinyuan, Sun, Ran, Shuai, Shuai, Pang, Shaoliang, Yang, Shiliang, Gao, Shuli, Yuan, Shanshan, Liu, Siqi, Deng, Shihong, Jiang, Shilei, Liu, Sitong, Cao, Tiancheng, Wang, Tianyu, Deng, Wenjin, Xie, Wuxun, Ming, Weipeng, He, Wenqing, Sun, Wen, Han, Xin, Huang, Xin, Deng, Xiaomin, Liu, Xiaojia, Wu, Xin, Zhao, Xu, Wei, Yanan, Yu, Yanbo, Cao, Yang, Li, Yangguang, Ma, Yangzhen, Xu, Yanming, Wang, Yaoyu, Shi, Yaqiang, Wang, Yilei, Zhou, Yizhuang, Zhong, Yinmin, Zhang, Yang, Wei, Yaoben, Luo, Yu, Lu, Yuanwei, Yin, Yuhe, Luo, Yuchu, Ding, Yuanhao, Yan, Yuting, Dai, Yaqi, Yang, Yuxiang, Xie, Zhe, Ge, Zheng, Sun, Zheng, Huang, Zhewei, Chang, Zhichao, Guan, Zhisheng, Yang, Zidong, Zhang, Zili, Jiao, Binxing, Jiang, Daxin, Shum, Heung-Yeung, Chen, Jiansheng, Li, Jing, Zhou, Shuchang, Zhang, Xiangyu, Zhang, Xinhao, and Zhu, Yibo

Subjects
Computer Science - Computation and Language, Computer Science - Artificial Intelligence, Computer Science - Human-Computer Interaction, Computer Science - Sound, Electrical Engineering and Systems Science - Audio and Speech Processing
Abstract

Real-time speech interaction, serving as a fundamental interface for human-machine collaboration, holds immense potential. However, current open-source models face limitations such as high costs in voice data collection, weakness in dynamic control, and limited intelligence. To address these challenges, this paper introduces Step-Audio, the first production-ready open-source solution. Key contributions include: 1) a 130B-parameter unified speech-text multi-modal model that achieves unified understanding and generation, with the Step-Audio-Chat version open-sourced; 2) a generative speech data engine that establishes an affordable voice cloning framework and produces the open-sourced lightweight Step-Audio-TTS-3B model through distillation; 3) an instruction-driven fine control system enabling dynamic adjustments across dialects, emotions, singing, and RAP; 4) an enhanced cognitive architecture augmented with tool calling and role-playing abilities to manage complex tasks effectively. Based on our new StepEval-Audio-360 evaluation benchmark, Step-Audio achieves state-of-the-art performance in human evaluations, especially in terms of instruction following. On open-source benchmarks like LLaMA Question, shows 9.3% average performance improvement, demonstrating our commitment to advancing the development of open-source multi-modal language technologies. Our code and models are available at https://github.com/stepfun-ai/Step-Audio.

Published
2025

2. Step-Video-T2V Technical Report: The Practice, Challenges, and Future of Video Foundation Model

Author

Ma, Guoqing, Huang, Haoyang, Yan, Kun, Chen, Liangyu, Duan, Nan, Yin, Shengming, Wan, Changyi, Ming, Ranchen, Song, Xiaoniu, Chen, Xing, Zhou, Yu, Sun, Deshan, Zhou, Deyu, Zhou, Jian, Tan, Kaijun, An, Kang, Chen, Mei, Ji, Wei, Wu, Qiling, Sun, Wen, Han, Xin, Wei, Yanan, Ge, Zheng, Li, Aojie, Wang, Bin, Huang, Bizhu, Wang, Bo, Li, Brian, Miao, Changxing, Xu, Chen, Wu, Chenfei, Yu, Chenguang, Shi, Dapeng, Hu, Dingyuan, Liu, Enle, Yu, Gang, Yang, Ge, Huang, Guanzhe, Yan, Gulin, Feng, Haiyang, Nie, Hao, Jia, Haonan, Hu, Hanpeng, Chen, Hanqi, Yan, Haolong, Wang, Heng, Guo, Hongcheng, Xiong, Huilin, Xiong, Huixin, Gong, Jiahao, Wu, Jianchang, Wu, Jiaoren, Wu, Jie, Yang, Jie, Liu, Jiashuai, Li, Jiashuo, Zhang, Jingyang, Guo, Junjing, Lin, Junzhe, Li, Kaixiang, Liu, Lei, Xia, Lei, Zhao, Liang, Tan, Liguo, Huang, Liwen, Shi, Liying, Li, Ming, Li, Mingliang, Cheng, Muhua, Wang, Na, Chen, Qiaohui, He, Qinglin, Liang, Qiuyan, Sun, Quan, Sun, Ran, Wang, Rui, Pang, Shaoliang, Yang, Shiliang, Liu, Sitong, Liu, Siqi, Gao, Shuli, Cao, Tiancheng, Wang, Tianyu, Ming, Weipeng, He, Wenqing, Zhao, Xu, Zhang, Xuelin, Zeng, Xianfang, Liu, Xiaojia, Yang, Xuan, Dai, Yaqi, Yu, Yanbo, Li, Yang, Deng, Yineng, Wang, Yingming, Wang, Yilei, Lu, Yuanwei, Chen, Yu, Luo, Yu, Luo, Yuchu, Yin, Yuhe, Feng, Yuheng, Yang, Yuxiang, Tang, Zecheng, Zhang, Zekai, Yang, Zidong, Jiao, Binxing, Chen, Jiansheng, Li, Jing, Zhou, Shuchang, Zhang, Xiangyu, Zhang, Xinhao, Zhu, Yibo, Shum, Heung-Yeung, and Jiang, Daxin

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Computation and Language
Abstract

We present Step-Video-T2V, a state-of-the-art text-to-video pre-trained model with 30B parameters and the ability to generate videos up to 204 frames in length. A deep compression Variational Autoencoder, Video-VAE, is designed for video generation tasks, achieving 16x16 spatial and 8x temporal compression ratios, while maintaining exceptional video reconstruction quality. User prompts are encoded using two bilingual text encoders to handle both English and Chinese. A DiT with 3D full attention is trained using Flow Matching and is employed to denoise input noise into latent frames. A video-based DPO approach, Video-DPO, is applied to reduce artifacts and improve the visual quality of the generated videos. We also detail our training strategies and share key observations and insights. Step-Video-T2V's performance is evaluated on a novel video generation benchmark, Step-Video-T2V-Eval, demonstrating its state-of-the-art text-to-video quality when compared with both open-source and commercial engines. Additionally, we discuss the limitations of current diffusion-based model paradigm and outline future directions for video foundation models. We make both Step-Video-T2V and Step-Video-T2V-Eval available at https://github.com/stepfun-ai/Step-Video-T2V. The online version can be accessed from https://yuewen.cn/videos as well. Our goal is to accelerate the innovation of video foundation models and empower video content creators., Comment: 36 pages, 14 figures

Published
2025

3. CoxKAN: Kolmogorov-Arnold Networks for Interpretable, High-Performance Survival Analysis

Author

Knottenbelt, William, Gao, Zeyu, Wray, Rebecca, Zhang, Woody Zhidong, Liu, Jiashuai, and Crispin-Ortuzar, Mireia

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

Survival analysis is a branch of statistics used for modeling the time until a specific event occurs and is widely used in medicine, engineering, finance, and many other fields. When choosing survival models, there is typically a trade-off between performance and interpretability, where the highest performance is achieved by black-box models based on deep learning. This is a major problem in fields such as medicine where practitioners are reluctant to blindly trust black-box models to make important patient decisions. Kolmogorov-Arnold Networks (KANs) were recently proposed as an interpretable and accurate alternative to multi-layer perceptrons (MLPs). We introduce CoxKAN, a Cox proportional hazards Kolmogorov-Arnold Network for interpretable, high-performance survival analysis. We evaluate the proposed CoxKAN on 4 synthetic datasets and 9 real medical datasets. The synthetic experiments demonstrate that CoxKAN accurately recovers interpretable symbolic formulae for the hazard function, and effectively performs automatic feature selection. Evaluation on the 9 real datasets show that CoxKAN consistently outperforms the Cox proportional hazards model and achieves performance that is superior or comparable to that of tuned MLPs. Furthermore, we find that CoxKAN identifies complex interactions between predictor variables that would be extremely difficult to recognise using existing survival methods, and automatically finds symbolic formulae which uncover the precise effect of important biomarkers on patient risk.

Published
2024

4. PAMIL: Prototype Attention-Based Multiple Instance Learning for Whole Slide Image Classification

Author

Liu, Jiashuai, Mao, Anyu, Niu, Yi, Zhang, Xianli, Gong, Tieliang, Li, Chen, Gao, Zeyu, Goos, Gerhard, Series Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Linguraru, Marius George, editor, Dou, Qi, editor, Feragen, Aasa, editor, Giannarou, Stamatia, editor, Glocker, Ben, editor, Lekadir, Karim, editor, and Schnabel, Julia A., editor

Published
2024
Full Text
View/download PDF

14. A Hybrid Open/Closed-Loop μ Control Method for Achieving Consistent Transient Performance in Turbofan Engines.

Author

Long, Yifu, Wang, Xi, and Liu, Jiashuai

Subjects
ACCELERATION (Mechanics), TURBOFAN engines, PRESSURE control, ALTITUDES, THRUST
Abstract

The inconsistency in acceleration and deceleration performance between high and low altitudes is a significant challenge in aircraft engine control today. In the past, neither open-loop fuel–air ratio control nor closed-loop N-dot control could resolve this issue perfectly; the difference in acceleration and deceleration performance between high and low altitudes is even more than three times. The operational characteristics of aircraft engines vary significantly between high and low altitudes, posing challenges for transient state control in high-performance aircraft engines. To address these transient performance inconsistencies due to altitude uncertainties, a μ-synthesis adaptive tracking transition control design method with hybrid open-loop and closed-loop direct thrust control is proposed. The main innovation lies in proposing a new segmented hybrid control scheme. Under a high-power state, it employs a dual closed-loop μ-synthesis adaptive tracking framework, using fuel flow to control thrust and nozzle area to control the turbine pressure ratio. In a low-power state, a single-variable closed-loop and open-loop control architecture is applied. Simulation results show that the hybrid open/closed-loop control method can suppress the inconsistency of acceleration and deceleration performance caused by altitude uncertainties in turbofan engines, ensuring consistent robustness in acceleration and deceleration performance across different altitudes. From the ground to an altitude of 11 km, the new method has an acceleration time range of 3.44 s–3.84 s and a deceleration time range of 4.83 s–5.98 s; compared with the previous fuel–air ratio acceleration time of 4.17 s–9.12 s and deceleration time of 6.12 s–14.48 s, its high and low-altitude acceleration and deceleration consistency performance is greatly improved. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

17. Dual-Loop μ-Synthesis Direct Thrust Control for Turbofan Engines.

Author

Long, Yifu, Wang, Xi, Zhao, Wenshuai, and Liu, Jiashuai

Subjects
THRUST, PRESSURE control, ROBUST control, THERMODYNAMICS, NOZZLES, TURBOFAN engines
Abstract

As the power unit of an aircraft, the engine's primary task is to provide the demanded thrust, making research on direct thrust control crucial. However, being a complicated multivariable system, effective multivariable direct thrust control methods are currently lacking. The main content of this paper is threefold. First, it presents a dual-loop multivariable μ-synthesis direct thrust control scheme for mixed-exhaust low-bypass turbofan engines, which is a typical rotationally symmetric machine. The scheme adjusts fuel flow for thrust control and nozzle area to control the turbine pressure ratio, ensuring thrust tracking while maintaining the engine's key parameters within safe limits. Second, a fast, accurate thrust estimation algorithm based on aerodynamic thermodynamics and component characteristics is introduced. At last, considering the model uncertainties between off-design and design points, a weight function frequency shaping μ-synthesis control design method is proposed to address internal loop coupling and external disturbance suppression. Nonlinear simulations within the flight envelope show that μ-synthesis direct thrust control achieves robust servo tracking and disturbance rejection, with a maximum steady-state thrust error of no more than 0.1%, and the key parameters are not over their safety boundaries. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

26. Feedforward compensation-based L1adaptive control for aeropropulsion system test facility and hardware-in-the-loop verification

Author

LIU, Jiashuai, HE, Ai, PEI, Xitong, and LONG, Yifu

Abstract

Aeropropulsion System Test Facility (ASTF) is required to accurately control the pressure and temperature of the airflow to test the performance of the aero-engine. However, the control accuracy of ASTF is significantly affected by the flow disturbance caused by aero-engine acceleration and deceleration. This would reduce the credibility of ASTF’s test results for the aero-engine. Therefore, first, this paper proposes a feedforward compensation-based L1adaptive control method for ASTF to address this problem. The baseline controller is first designed based on ideal uncoupled closed-loop dynamics to achieve dynamic decoupling. Then, L1adaptive control is adopted to deal with various uncertainties and ensure good control performance. To further enhance the anti-disturbance performance, a feedforward strategy based on disturbance prediction is designed in the L1adaptive control framework to compensate for the unmatched flow disturbance, which cannot be measured directly. In addition, this strategy takes into account the effects of actuator dynamics. With this method, the feedforward term can be determined from the nominal model parameters despite uncertainties. Finally, to demonstrate the effectiveness of the proposed method, various comparative experiments are performed on a hardware-in-the-loop system of ASTF. The experimental results show that the proposed method possesses excellent tracking performance, anti-disturbance performance and robustness.

Published
2025
Full Text
View/download PDF

32. Effect of exercise rehabilitation on hemodynamic performance after carotid artery stenting: a numerical study.

Author

Liu J, Fan Z, Liu X, Xu X, Liu M, Ye X, and Deng X

Subjects
Humans, Hemodynamics, Carotid Arteries surgery, Models, Cardiovascular, Stents, Carotid Stenosis surgery
Abstract

A high in-stent restenosis rate and thrombosis have compromised clinical benefits after vascular stent placement. Exercise rehabilitation after stenting emerges as a promising and practical therapeutic strategy to improve the clinical performance of this therapy, although it remains controversial. The present study aimed to explore the impact of exercise training on hemodynamic performance after vascular stent implantation. Different 3-dimensional computational models based on the patient-specific carotids were constructed to calculate hemodynamic parameters, including flow velocity, time-averaged wall shear (TAWSS), oscillatory shear index (OSI) and relative residence time (RRT). The results demonstrated that exercise training increased TAWSS but decreased OSI and RRT in some cases after the intervention, and high-intensity exercise further suppressed the adverse blood flow. However, exercise training remarkably reduced TAWSS and elevated OSI and RRT in patients with mild stenosis at upstream of stented segment. Additionally, we discovered that the hemodynamic environment change induced by exercise training was not significant compared to the stent position in some cases. Exercise had a less beneficial impact on the disturbed blood flow after the distal common carotid artery (CCA) stenting. These findings highlighted that exercise-induced hemodynamic changes differ under different conditions. The exercise training for the intervention patients should only be performed after a comprehensive vascular function assessment.

Published
2023

Catalog

Books, media, physical & digital resources
See catalog results