Search

Your search keyword '"Li, Zhenni"' showing total 131 results
Start Over Author "Li, Zhenni" Search Limiters Available in Library Collection
131 results on '"Li, Zhenni"'

1. Compressing Deep Reinforcement Learning Networks with a Dynamic Structured Pruning Method for Autonomous Driving

Author

Su, Wensheng, Li, Zhenni, Xu, Minrui, Kang, Jiawen, Niyato, Dusit, and Xie, Shengli

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

Deep reinforcement learning (DRL) has shown remarkable success in complex autonomous driving scenarios. However, DRL models inevitably bring high memory consumption and computation, which hinders their wide deployment in resource-limited autonomous driving devices. Structured Pruning has been recognized as a useful method to compress and accelerate DRL models, but it is still challenging to estimate the contribution of a parameter (i.e., neuron) to DRL models. In this paper, we introduce a novel dynamic structured pruning approach that gradually removes a DRL model's unimportant neurons during the training stage. Our method consists of two steps, i.e. training DRL models with a group sparse regularizer and removing unimportant neurons with a dynamic pruning threshold. To efficiently train the DRL model with a small number of important neurons, we employ a neuron-importance group sparse regularizer. In contrast to conventional regularizers, this regularizer imposes a penalty on redundant groups of neurons that do not significantly influence the output of the DRL model. Furthermore, we design a novel structured pruning strategy to dynamically determine the pruning threshold and gradually remove unimportant neurons with a binary mask. Therefore, our method can remove not only redundant groups of neurons of the DRL model but also achieve high and robust performance. Experimental results show that the proposed method is competitive with existing DRL pruning methods on discrete control environments (i.e., CartPole-v1 and LunarLander-v2) and MuJoCo continuous environments (i.e., Hopper-v3 and Walker2D-v3). Specifically, our method effectively compresses $93\%$ neurons and $96\%$ weights of the DRL model in four challenging DRL environments with slight accuracy degradation.

Published
2024

12. Adaptive Output Synchronization With Designated Convergence Rate of Multiagent Systems Based on Off-Policy Reinforcement Learning

Author

Huang, Chengjie, Chen, Ci, Xie, Kan, Li, Zhenni, and Xie, Shengli

Abstract

In this article, an optimal output synchronization solution to the $H_{\infty }$ optimization of linear discrete-time (DT) multiagent systems is investigated. Compared with current approaches, the issue of designated convergence rate is handled with system optimality, while less computation cost is required. Specifically, the internal model principle is employed to derive a cooperative regulation problem of DT systems, wherein no explicit solution to output regulation equations is needed for learning. Then, we introduce a convergence rate parameter to construct a group of auxiliary cooperative systems, based on which the zero-sum game in $H_{\infty }$ optimization is formulated. The data-efficient off-policy reinforcement learning and output-feedback technique are applied to solve the enhanced Bellman equations with a designated convergence rate. This results in an online optimal synchronization solution learning from only the input-output data along the system trajectories. It is shown that the proposed optimal synchronization protocol achieves asymptotic synchronization for the original systems with the consensus error converging to zero at a designated rate. The effectiveness of the proposed approach is verified by the simulation results.

Published
2024
Full Text
View/download PDF

20. A Few-Shot Learning-Based Semi-Supervised Method to Enhance NLOS Recognition Accuracy Across Multiple Locations by Cross-Spatiotemporal Information

Author

Zeng, Kungan, Li, Zhenni, Xie, Shengli, Wang, Qianming, Chen, Wuhui, and Zheng, Zibin

Abstract

Machine learning (ML)-based Global Navigation Satellite System (GNSS) signal recognition has been considered an efficient way to mitigate multipath error in positioning estimation. However, because of the spatial heterogeneity of nonline-of-sight (NLOS) interference, existing NLOS signal recognition methods struggle to achieve satisfactory accuracy across multiple locations. In particular, current methods achieve limited recognition accuracy on target domains containing unfamiliar environments. This study proposes a few-shot learning-based semi-supervised GNSS NLOS recognition method using cross-spatiotemporal information, which can achieve high-precision recognition for multiple target domains containing unfamiliar environments only requiring a limited number of labeled samples. First, we propose a self-attention mechanism-based cross-spatiotemporal information (SAM-CSI) module that captures the environmental characteristics used to model the spatial heterogeneity of NLOS interference. Next, to avoid the need for numerous labeled samples in training the model, we introduce a Siamese neural network-based few-shot learning (SNN-FSL) architecture to perform NLOS recognition using a similarity metric instead of label prediction. We then develop a semi-supervised NLOS recognition method by integrating the SAM-CSI module with the SNN-FSL architecture to enable high-precision recognition in target domains containing unfamiliar environments. Finally, we conducted comparison and few-shot learning experiments using real-world datasets, which demonstrated that our proposed method could achieve over 90% recognition accuracy in target domains containing unfamiliar environments after 50-shot learning. It represented an improvement of over 19% compared with existing methods.

Published
2024
Full Text
View/download PDF

21. Dictionary Learning-Structured Reinforcement Learning With Adaptive-Sparsity Regularizer

Author

Li, Zhenni, Tang, Jianhao, Zhao, Haoli, Chen, Ci, and Xie, Shengli

Abstract

Deep reinforcement learning (DRL) has been applied to satellite navigation and positioning applications. Its performance relies heavily on the function-approximation capability of deep neural networks. However, existing DRL models suffer from catastrophic interference, resulting in inaccurate function approximation. The sparse-coding-based DRL is an effective method to mitigating this interference, but existing methods involve the following two challenging issues: first, the value function estimation network suffers from instability problems with gradient backpropagation, including gradient explosion and gradient vanishing, second, existing methods are limited to using hand-crafted sparse regularizers that produce only static sparsity, which may be difficult to apply in various dynamic reinforcement learning (RL) environments. In this article, we propose a novel dictionary learning (DL)-structured RL model with adaptive-sparsity regularizer (ASR) that alleviates the catastrophic interference and enables accurate value function approximation, thereby improving the RL performance. To alleviate the interference and avoid the instability problems in RL, a feedforward DL-structured RL model is constructed to predict the value function without the need for gradient backpropagation. To learn data-driven sparse representations with adaptive sparsity, we propose to use the learnable sparse regularizer ASR in the model, where the key hyperparameters of ASR can be trained to be adaptive to variable RL environments. To optimize the model efficiently, the model parameters are first pretrained in the pretraining stage, with only the value weights used for value function approximation needing to be fine-tuned for actual RL applications in the control training stage. Our comparative experiments in benchmark environments demonstrate that the proposed method can outperform existing state-of-the-art sparse-coding-based RL algorithms. In terms of accumulated rewards (used to measure the quality of the learned policy), the improvement was over 63% in Cart Pole environment and nearly 10% for Puddle World. Furthermore, the proposed algorithm can maintain its relatively high performance in the presence of noise up to 20 dB.

Published
2024
Full Text
View/download PDF

36. Spot Detection for Laser Sensors Based on Annular Convolution Filtering.

Author

Li, Lingjiang, Li, Maolin, Sun, Weijun, Li, Zhenni, and Yang, Zuyuan

Subjects
LASER based sensors, OPTICAL interference, SPECIAL effects in lighting
Abstract

Spot detection has attracted continuous attention for laser sensors with applications in communication, measurement, etc. The existing methods often directly perform binarization processing on the original spot image. They suffer from the interference of the background light. To reduce this kind of interference, we propose a novel method called annular convolution filtering (ACF). In our method, the region of interest (ROI) in the spot image is first searched by using the statistical properties of pixels. Then, the annular convolution strip is constructed based on the energy attenuation property of the laser and the convolution operation is performed in the ROI of the spot image. Finally, a feature similarity index is designed to estimate the parameters of the laser spot. Experiments on three datasets with different kinds of background light show the advantages of our ACF method, with comparison to the theoretical method based on international standard, the practical method used in the market products, and the recent benchmark methods AAMED and ALS. [ABSTRACT FROM AUTHOR]

Published
2023
Full Text
View/download PDF

39. Optimal BR signalling is required for adequate cell wall orientation in the Arabidopsis root meristem

Author

Li, Zhenni, Sela, Ayala, Fridman, Yulia, Garstka, Lucía, Höfte, Herman, Savaldi-Goldstein, Sigal, and Wolf, Sebastian

Subjects
Cell division, Arabidopsis Proteins, Cell wall, fungi, Meristem, Arabidopsis, Plant Development, Plant Roots, Root, Brassinosteroids, Homeostasis, Pectins, Research Article, Cytokinesis, Signal Transduction
Abstract

Plant brassinosteroid hormones (BRs) regulate growth in part through altering the properties of the cell wall, the extracellular matrix of plant cells. Conversely, feedback signalling from the wall connects the state of cell wall homeostasis to the BR receptor complex and modulates BR activity. Here, we report that both pectin-triggered cell wall signalling and impaired BR signalling result in altered cell wall orientation in the Arabidopsis root meristem. Furthermore, both depletion of endogenous BRs and exogenous supply of BRs triggered these defects. Cell wall signalling-induced alterations in the orientation of newly placed walls appear to occur late during cytokinesis, after initial positioning of the cortical division zone. Tissue-specific perturbations of BR signalling revealed that the cellular malfunction is unrelated to previously described whole organ growth defects. Thus, tissue type separates the pleiotropic effects of cell wall/BR signals and highlights their importance during cell wall placement., Summary: Both increased and reduced BR signalling strength results in altered cell wall orientation in the Arabidopsis root, uncoupled from whole-root growth control.

Published
2021

41. A Randomized, Double-Blind, Single-Dose Study Comparing the Biosimilarity of HOT-1010 With Bevacizumab (Avastin®) in Chinese Healthy Male Subjects

Author

Huang, Kai, primary, Que, Linling, additional, Ding, Ying, additional, Chu, Nannan, additional, Qian, Zhenzhong, additional, Shi, Yunfei, additional, Qin, Wei, additional, Li, Zhenni, additional, Chen, Yuanxin, additional, Gu, Xianghong, additional, Wang, Jiakun, additional, Zhang, Lin, additional, Zhang, Jisheng, additional, Zhu, Xiangyang, additional, Yang, Yongmin, additional, Tang, Yuan, additional, and He, Qing, additional

Published
2021
Full Text
View/download PDF

44. Nisin-loaded polydopamine/hydroxyapatite composites

Author

Wang, Ni, Yu, Xing, Kong, Qingjun, Li, Zhenni, Li, Penghui, Ren, Xueyan, Peng, Bo, Deng, Ziwei, Shaanxi Normal University, Shenzhen Institute of Advanced Technology, Department of Applied Physics, Aalto-yliopisto, and Aalto University

Subjects
Polydopamine, Biomimetic mineralization, stomatognathic system, education, Antibacterial activity, Nisin, Hydroxyapatite
Abstract

Microbial contamination caused by bacterial adhesion and colonization onto the materials surfaces has long been a serious threat to human health. Thus, it is imperative to develop advanced antimicrobial materials for bio-applications. Here, a unique polydopamine/ hydroxyapatite/nisin (PDA/HAP/Nisin) composite with good bioactivities and enhanced antibacterial activities have been developed. The uniform PDA particles are first prepared via the dopamine self-polymerization in an alkaline water-ethanol mixed solvent. Due to the abundance surface catecholamine moieties, PDA particles can serve as the powerful templates for the in-situ mineralization of hydroxyapatite (HAP) in a simulated body fluid (SBF) solution, forming PDA/HAP composites. Transmission electron microscopy, scanning electron microscopy, transform infrared spectroscopy, energy-dispersive X-ray spectroscopy, X-ray photoelectron spectroscopy, and X-ray diffraction are used to investigate the morphology, surface composition, and crystalline structure of the PDA/HAP composites. These PDA/HAP composites are utilized as the reservoir for loading the antimicrobial peptide–nisin. Ultraviolet visible spectrophotometry analysis shows that PDA/HAP composites have exhibited a high loading capacity and a superior loading efficiency for nisin. The resulting PDA/HAP/Nisin composites are ideal antimicrobial biomaterials for bio-application because they exhibit not only a significant in vitro cytocompatibility to human embryonic kidney HEK293 T cells, but also the enhanced antibacterial activities against Staphylococcus aureus (Gram-positive bacteria). These advantageous properties of the PDA/HAP/Nisin composites allow them as the promising candidates for a broad range of bio-applications, e.g., drug delivery, implant engineering, and bio-scaffold.

Published
2020

48. The role of cell identity in the response to cell wall perturbation in the Arabidopsis thaliana primary root

Author

Li, Zhenni

Subjects
food and beverages, 570 Life sciences
Abstract

The cell wall is a defining feature of plant cells. It is a rigid, yet flexible, layer surrounding each cell outside the plasma membrane and is mainly composed of polysaccharides such as cellulose, hemicellulose, and pectin, as well asm to a lesser extent, phenolic compounds such as lignin and structural proteins. Plant cell wall is central to nearly all aspects of plant. Its biosynthesis and remodeling are essential for cell division, expansion and differentiation, the fundamental events through plant growth and development. In brief, cell walls shape the cells, which is crucial for organ formation, provide mechanical support to the plant, control cell-to-cell adhesion, form the interface between cells, which is indispensable for cellular communication, and regulate plant-pathogen/environment interactions. As plants permanently face intrinsic and extrinsic cues from developmental programs and the environment, this information must be correctly conveyed into the cells in order to adjust the plant’s growth behavior and reallocate the resources accordingly. Therefore, a cell wall surveillance and signalling system must exist to ensure transmission of the information. Roots form the hidden half of the plant and perform numerous physical and physiological functions such as anchoring, nutrient uptake and transport. The primary root of the model organism in plant research, Arabidopsis thaliana, is formed with different tissues organized as longitudinal cell files that are radially patterned in a concentric manner, which provides an easily accessible model to study plant development. By using this model, previous studies in our lab have revealed the plasma membrane residing RECEPTOR LIKE PROTEIN (RLP)44 as key factor for cell wall surveillance and signalling. RLP44 monitors pectin status in the cell wall and transduces the signal of changes in cell wall integrity through interaction with the brassinosteroid receptor BRI1, which in turn activates a well-described signalling cascade to regulate plant growth transcriptionally. As in plants, different tissues fulfill distinct biological functions and have similar, yet distinguishable, characteristics in their cell walls, investigation of RLP44-mediated cell wall signalling in tissue-specific context would bring more insights into how plants co-ordinates its growth and development in responses to cell wall changes at tissue level. To do so, we made use of the recently developed GreenGate cloning technique and the dexamethasone-inducible system pOp6/GR-LhG4. By using promoters driving expression specifically in different cell types/ tissues, we ectopically expressed pectin-modifying gene PECTIN METHYLESTERASE INHIBITOR 5 (PMEI5) and studied the responses of the plant to the loss of cell wall integrity. Here, we showed that cell wall homeostasis had pronounced impact on root growth, especially on in root morphogenesis, meristem size control, division plane determination through CDS maintenance, and tissue patterning. This influence was manifested as varying phenotypes at tissue level, implemented probably through crosstalks between brassinosteroid and other hormone signalling pathways in both cell-autonomous and non cell-autonomous manner. Besides, the previously described RLP44-mediated cell wall signalling exerts its role in a cell type/tissue-specific way and other cell wall sensing and signalling mechanisms are also possibly involved in the responses to cell wall perturbation. However, biochemical and histological characterization of cell wall properties as well as genetic studies for discovering the underlying molecular mechanisms will be necessary to further understand the role of cell identity in cell wall signalling and plant growth regulation.

Published
2018
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results