25,291 results on '"Fu, Qiang"'
Search Results
152. Correction to: An Innovative Mechanism of Directional Rock Cracking Mechanics and Mine Pressure Control Method for Energy-Gathering Blasting
- Author
-
Xu, Xiaoding, Zhou, Yuejin, Yang, Jun, Gao, Yubing, Zhu, Chun, Wang, Yajun, and Fu, Qiang
- Published
- 2024
- Full Text
- View/download PDF
153. Author Correction: PDK4-dependent hypercatabolism and lactate production of senescent cells promotes cancer malignancy
- Author
-
Dou, Xuefeng, Fu, Qiang, Long, Qilai, Liu, Shuning, Zou, Yejun, Fu, Da, Xu, Qixia, Jiang, Zhirui, Ren, Xiaohui, Zhang, Guilong, Wei, Xiaoling, Li, Qingfeng, Campisi, Judith, Zhao, Yuzheng, and Sun, Yu
- Published
- 2024
- Full Text
- View/download PDF
154. RLogist: Fast Observation Strategy on Whole-slide Images with Deep Reinforcement Learning
- Author
-
Zhao, Boxuan, Zhang, Jun, Ye, Deheng, Cao, Jian, Han, Xiao, Fu, Qiang, and Yang, Wei
- Subjects
Computer Science - Computer Vision and Pattern Recognition ,Computer Science - Artificial Intelligence ,Computer Science - Machine Learning - Abstract
Whole-slide images (WSI) in computational pathology have high resolution with gigapixel size, but are generally with sparse regions of interest, which leads to weak diagnostic relevance and data inefficiency for each area in the slide. Most of the existing methods rely on a multiple instance learning framework that requires densely sampling local patches at high magnification. The limitation is evident in the application stage as the heavy computation for extracting patch-level features is inevitable. In this paper, we develop RLogist, a benchmarking deep reinforcement learning (DRL) method for fast observation strategy on WSIs. Imitating the diagnostic logic of human pathologists, our RL agent learns how to find regions of observation value and obtain representative features across multiple resolution levels, without having to analyze each part of the WSI at the high magnification. We benchmark our method on two whole-slide level classification tasks, including detection of metastases in WSIs of lymph node sections, and subtyping of lung cancer. Experimental results demonstrate that RLogist achieves competitive classification performance compared to typical multiple instance learning algorithms, while having a significantly short observation path. In addition, the observation path given by RLogist provides good decision-making interpretability, and its ability of reading path navigation can potentially be used by pathologists for educational/assistive purposes. Our code is available at: \url{https://github.com/tencent-ailab/RLogist}., Comment: accepted by AAAI 2023
- Published
- 2022
155. DGI: Easy and Efficient Inference for GNNs
- Author
-
Yin, Peiqi, Yan, Xiao, Zhou, Jinjing, Fu, Qiang, Cai, Zhenkun, Cheng, James, Tang, Bo, and Wang, Minjie
- Subjects
Computer Science - Machine Learning - Abstract
While many systems have been developed to train Graph Neural Networks (GNNs), efficient model inference and evaluation remain to be addressed. For instance, using the widely adopted node-wise approach, model evaluation can account for up to 94% of the time in the end-to-end training process due to neighbor explosion, which means that a node accesses its multi-hop neighbors. On the other hand, layer-wise inference avoids the neighbor explosion problem by conducting inference layer by layer such that the nodes only need their one-hop neighbors in each layer. However, implementing layer-wise inference requires substantial engineering efforts because users need to manually decompose a GNN model into layers for computation and split workload into batches to fit into device memory. In this paper, we develop Deep Graph Inference (DGI) -- a system for easy and efficient GNN model inference, which automatically translates the training code of a GNN model for layer-wise execution. DGI is general for various GNN models and different kinds of inference requests, and supports out-of-core execution on large graphs that cannot fit in CPU memory. Experimental results show that DGI consistently outperforms layer-wise inference across different datasets and hardware settings, and the speedup can be over 1,000x., Comment: 10 pages, 10 figures
- Published
- 2022
156. DIGMN: Dynamic Intent Guided Meta Network for Differentiated User Engagement Forecasting in Online Professional Social Platforms
- Author
-
Li, Feifan, Du, Lun, Fu, Qiang, Han, Shi, Du, Yushu, Lu, Guangming, and Li, Zi
- Subjects
Computer Science - Machine Learning - Abstract
User engagement prediction plays a critical role for designing interaction strategies to grow user engagement and increase revenue in online social platforms. Through the in-depth analysis of the real-world data from the world's largest professional social platforms, i.e., LinkedIn, we find that users expose diverse engagement patterns, and a major reason for the differences in user engagement patterns is that users have different intents. That is, people have different intents when using LinkedIn, e.g., applying for jobs, building connections, or checking notifications, which shows quite different engagement patterns. Meanwhile, user intents and the corresponding engagement patterns may change over time. Although such pattern differences and dynamics are essential for user engagement prediction, differentiating user engagement patterns based on user dynamic intents for better user engagement forecasting has not received enough attention in previous works. In this paper, we proposed a Dynamic Intent Guided Meta Network (DIGMN), which can explicitly model user intent varying with time and perform differentiated user engagement forecasting. Specifically, we derive some interpretable basic user intents as prior knowledge from data mining and introduce prior intents in explicitly modeling dynamic user intent. Furthermore, based on the dynamic user intent representations, we propose a meta predictor to perform differentiated user engagement forecasting. Through a comprehensive evaluation on LinkedIn anonymous user data, our method outperforms state-of-the-art baselines significantly, i.e., 2.96% and 3.48% absolute error reduction, on coarse-grained and fine-grained user engagement prediction tasks, respectively, demonstrating the effectiveness of our method., Comment: 10 pages, Accepted by WSDM'23
- Published
- 2022
157. Minimum distances of binary optimal LCD codes of dimension five are completely determined
- Author
-
Liu, Yang, Li, Ruihu, Fu, Qiang, and Song, Hao
- Subjects
Computer Science - Information Theory - Abstract
Let $t \in \{2,8,10,12,14,16,18\}$ and $n=31s+t\geq 14$, $d_{a}(n,5)$ and $d_{l}(n,5)$ be distances of binary $[n,5]$ optimal linear codes and optimal linear complementary dual (LCD) codes, respectively. We show that an $[n,5,d_{a}(n,5)]$ optimal linear code is not an LCD code, there is an $[n,5,d_{l}(n,5)]=[n,5,d_{a}(n,5)-1]$ optimal LCD code if $t\neq 16$, and an optimal $[n,5,d_{l}(n,5)]$ optimal LCD code has $d_{l}(n,5)=16s+6=d_{a}(n,5)-2$ for $t=16$. Combined with known results on optimal LCD code, $d_{l}(n,5)$ of all $[n,5]$ LCD codes are completely determined., Comment: 15pages,7tables
- Published
- 2022
158. Revisiting Discrete Soft Actor-Critic
- Author
-
Zhou, Haibin, Lin, Zichuan, Li, Junyou, Fu, Qiang, Yang, Wei, and Ye, Deheng
- Subjects
Computer Science - Machine Learning ,Computer Science - Artificial Intelligence - Abstract
We study the adaption of soft actor-critic (SAC) from continuous action space to discrete action space. We revisit vanilla SAC and provide an in-depth understanding of its Q value underestimation and performance instability issues when applied to discrete settings. We thereby propose entropy-penalty and double average Q-learning with Q-clip to address these issues. Extensive experiments on typical benchmarks with discrete action space, including Atari games and a large-scale MOBA game, show the efficacy of our proposed method. Our code is at:https://github.com/coldsummerday/Revisiting-Discrete-SAC.
- Published
- 2022
159. Honor of Kings Arena: an Environment for Generalization in Competitive Reinforcement Learning
- Author
-
Wei, Hua, Chen, Jingxiao, Ji, Xiyang, Qin, Hongyang, Deng, Minwen, Li, Siqin, Wang, Liang, Zhang, Weinan, Yu, Yong, Liu, Lin, Huang, Lanxiao, Ye, Deheng, Fu, Qiang, and Yang, Wei
- Subjects
Computer Science - Machine Learning ,Computer Science - Artificial Intelligence - Abstract
This paper introduces Honor of Kings Arena, a reinforcement learning (RL) environment based on Honor of Kings, one of the world's most popular games at present. Compared to other environments studied in most previous work, ours presents new generalization challenges for competitive reinforcement learning. It is a multi-agent problem with one agent competing against its opponent; and it requires the generalization ability as it has diverse targets to control and diverse opponents to compete with. We describe the observation, action, and reward specifications for the Honor of Kings domain and provide an open-source Python-based interface for communicating with the game engine. We provide twenty target heroes with a variety of tasks in Honor of Kings Arena and present initial baseline results for RL-based methods with feasible computing resources. Finally, we showcase the generalization challenges imposed by Honor of Kings Arena and possible remedies to the challenges. All of the software, including the environment-class, are publicly available at https://github.com/tencent-ailab/hok_env . The documentation is available at https://aiarena.tencent.com/hok/doc/ ., Comment: Accepted by NeurIPS 2022
- Published
- 2022
160. Make Heterophily Graphs Better Fit GNN: A Graph Rewiring Approach
- Author
-
Bi, Wendong, Du, Lun, Fu, Qiang, Wang, Yanlin, Han, Shi, and Zhang, Dongmei
- Subjects
Computer Science - Machine Learning ,Computer Science - Artificial Intelligence - Abstract
Graph Neural Networks (GNNs) are popular machine learning methods for modeling graph data. A lot of GNNs perform well on homophily graphs while having unsatisfactory performance on heterophily graphs. Recently, some researchers turn their attention to designing GNNs for heterophily graphs by adjusting the message passing mechanism or enlarging the receptive field of the message passing. Different from existing works that mitigate the issues of heterophily from model design perspective, we propose to study heterophily graphs from an orthogonal perspective by rewiring the graph structure to reduce heterophily and making the traditional GNNs perform better. Through comprehensive empirical studies and analysis, we verify the potential of the rewiring methods. To fully exploit its potential, we propose a method named Deep Heterophily Graph Rewiring (DHGR) to rewire graphs by adding homophilic edges and pruning heterophilic edges. The detailed way of rewiring is determined by comparing the similarity of label/feature-distribution of node neighbors. Besides, we design a scalable implementation for DHGR to guarantee high efficiency. DHRG can be easily used as a plug-in module, i.e., a graph pre-processing step, for any GNNs, including both GNN for homophily and heterophily, to boost their performance on the node classification task. To the best of our knowledge, it is the first work studying graph rewiring for heterophily graphs. Extensive experiments on 11 public graph datasets demonstrate the superiority of our proposed methods., Comment: 11 pages
- Published
- 2022
161. Dynamics-Adaptive Continual Reinforcement Learning via Progressive Contextualization
- Author
-
Zhang, Tiantian, Lin, Zichuan, Wang, Yuxing, Ye, Deheng, Fu, Qiang, Yang, Wei, Wang, Xueqian, Liang, Bin, Yuan, Bo, and Li, Xiu
- Subjects
Computer Science - Machine Learning ,Computer Science - Artificial Intelligence - Abstract
A key challenge of continual reinforcement learning (CRL) in dynamic environments is to promptly adapt the RL agent's behavior as the environment changes over its lifetime, while minimizing the catastrophic forgetting of the learned information. To address this challenge, in this article, we propose DaCoRL, i.e., dynamics-adaptive continual RL. DaCoRL learns a context-conditioned policy using progressive contextualization, which incrementally clusters a stream of stationary tasks in the dynamic environment into a series of contexts and opts for an expandable multihead neural network to approximate the policy. Specifically, we define a set of tasks with similar dynamics as an environmental context and formalize context inference as a procedure of online Bayesian infinite Gaussian mixture clustering on environment features, resorting to online Bayesian inference to infer the posterior distribution over contexts. Under the assumption of a Chinese restaurant process prior, this technique can accurately classify the current task as a previously seen context or instantiate a new context as needed without relying on any external indicator to signal environmental changes in advance. Furthermore, we employ an expandable multihead neural network whose output layer is synchronously expanded with the newly instantiated context, and a knowledge distillation regularization term for retaining the performance on learned tasks. As a general framework that can be coupled with various deep RL algorithms, DaCoRL features consistent superiority over existing methods in terms of the stability, overall performance and generalization ability, as verified by extensive experiments on several robot navigation and MuJoCo locomotion tasks., Comment: Accepted by IEEE Transactions on Neural Networks and Learning Systems, 2023
- Published
- 2022
162. Learning Rate Perturbation: A Generic Plugin of Learning Rate Schedule towards Flatter Local Minima
- Author
-
Liu, Hengyu, Fu, Qiang, Du, Lun, Zhang, Tiancheng, Yu, Ge, Han, Shi, and Zhang, Dongmei
- Subjects
Computer Science - Machine Learning - Abstract
Learning rate is one of the most important hyper-parameters that has a significant influence on neural network training. Learning rate schedules are widely used in real practice to adjust the learning rate according to pre-defined schedules for fast convergence and good generalization. However, existing learning rate schedules are all heuristic algorithms and lack theoretical support. Therefore, people usually choose the learning rate schedules through multiple ad-hoc trials, and the obtained learning rate schedules are sub-optimal. To boost the performance of the obtained sub-optimal learning rate schedule, we propose a generic learning rate schedule plugin, called LEArning Rate Perturbation (LEAP), which can be applied to various learning rate schedules to improve the model training by introducing a certain perturbation to the learning rate. We found that, with such a simple yet effective strategy, training processing exponentially favors flat minima rather than sharp minima with guaranteed convergence, which leads to better generalization ability. In addition, we conduct extensive experiments which show that training with LEAP can improve the performance of various deep learning models on diverse datasets using various learning rate schedules (including constant learning rate)., Comment: 7 pages, Accepted by CIKM'22
- Published
- 2022
163. MM-GNN: Mix-Moment Graph Neural Network towards Modeling Neighborhood Feature Distribution
- Author
-
Bi, Wendong, Du, Lun, Fu, Qiang, Wang, Yanlin, Han, Shi, and Zhang, Dongmei
- Subjects
Computer Science - Machine Learning - Abstract
Graph Neural Networks (GNNs) have shown expressive performance on graph representation learning by aggregating information from neighbors. Recently, some studies have discussed the importance of modeling neighborhood distribution on the graph. However, most existing GNNs aggregate neighbors' features through single statistic (e.g., mean, max, sum), which loses the information related to neighbor's feature distribution and therefore degrades the model performance. In this paper, inspired by the method of moment in statistical theory, we propose to model neighbor's feature distribution with multi-order moments. We design a novel GNN model, namely Mix-Moment Graph Neural Network (MM-GNN), which includes a Multi-order Moment Embedding (MME) module and an Element-wise Attention-based Moment Adaptor module. MM-GNN first calculates the multi-order moments of the neighbors for each node as signatures, and then use an Element-wise Attention-based Moment Adaptor to assign larger weights to important moments for each node and update node representations. We conduct extensive experiments on 15 real-world graphs (including social networks, citation networks and web-page networks etc.) to evaluate our model, and the results demonstrate the superiority of MM-GNN over existing state-of-the-art models., Comment: accepted by WSDM2023
- Published
- 2022
- Full Text
- View/download PDF
164. Heterogeneous Multi-agent Zero-Shot Coordination by Coevolution
- Author
-
Xue, Ke, Wang, Yutong, Guan, Cong, Yuan, Lei, Fu, Haobo, Fu, Qiang, Qian, Chao, and Yu, Yang
- Subjects
Computer Science - Neural and Evolutionary Computing ,Computer Science - Artificial Intelligence ,Computer Science - Machine Learning ,Computer Science - Multiagent Systems - Abstract
Generating agents that can achieve zero-shot coordination (ZSC) with unseen partners is a new challenge in cooperative multi-agent reinforcement learning (MARL). Recently, some studies have made progress in ZSC by exposing the agents to diverse partners during the training process. They usually involve self-play when training the partners, implicitly assuming that the tasks are homogeneous. However, many real-world tasks are heterogeneous, and hence previous methods may be inefficient. In this paper, we study the heterogeneous ZSC problem for the first time and propose a general method based on coevolution, which coevolves two populations of agents and partners through three sub-processes: pairing, updating and selection. Experimental results on various heterogeneous tasks highlight the necessity of considering the heterogeneous setting and demonstrate that our proposed method is a promising solution for heterogeneous ZSC tasks.
- Published
- 2022
165. Various Wavefront Sensing and Control Developments on the Santa Cruz Extreme AO Laboratory (SEAL) Testbed
- Author
-
Gerard, Benjamin L., Perez-Soto, Javier, Chambouleyron, Vincent, van Kooten, Maaike A. M., Dillon, Daren, Cetre, Sylvain, Jensen-Clem, Rebecca, Fu, Qiang, Amata, Hadi, and Heidrich, Wolfgang
- Subjects
Astrophysics - Instrumentation and Methods for Astrophysics ,Astrophysics - Earth and Planetary Astrophysics - Abstract
Ground-based high contrast imaging (HCI) and extreme adaptive optics (AO) technologies have advanced to the point of enabling direct detections of gas-giant exoplanets orbiting beyond the snow lines around nearby young star systems. However, leftover wavefront errors using current HCI and AO technologies, realized as "speckles" in the coronagraphic science image, still limit HCI instrument sensitivities to detecting and characterizing lower-mass, closer-in, and/or older/colder exoplanetary systems. Improving the performance of AO wavefront sensors (WFSs) and control techniques is critical to improving such HCI instrument sensitivity. Here we present three different ongoing wavefront sensing and control project developments on the Santa cruz Extreme AO Laboratory (SEAL) testbed: (1) "multi-WFS single congugate AO (SCAO)" using the Fast Atmospheric Self-coherent camera (SCC) Technique (FAST) and a Shack Hartmann WFS, (2) pupil chopping for focal plane wavefront sensing, first with an external amplitude modulator and then with the DM as a phase-only modulator, and (3) a laboratory demonstration of enhanced linearity with the non-modulated bright Pyramid WFS (PWFS) compared to the regular PWFS. All three topics share a common theme of multi-WFS SCAO and/or second stage AO, presenting opportunities and applications to further investigate these techniques in the future., Comment: submitted to SPIE Astronomical Telescopes and Instrumentation 2022, paper number 12185-89
- Published
- 2022
166. One-step exfoliation method for plasmonic activation of large-area 2D crystals
- Author
-
Fu, Qiang, Dai, Jia-Qi, Huang, Xin-Yu, Dai, Yun-Yun, Pan, Yu-Hao, Yang, Long-Long, Sun, Zhen-Yu, Miao, Tai-Min, Zhou, Meng-Fan, Zhao, Lin, Zhao, Wei-Jie, Han, Xu, Lu, Jun-Peng, Gao, Hong-Jun, Zhou, Xing-Jiang, Wang, Ye-Liang, Ni, Zhen-Hua, Ji, Wei, and Huang, Yuan
- Subjects
Condensed Matter - Materials Science ,Physics - Optics - Abstract
Advanced exfoliation techniques are crucial for exploring the intrinsic properties and applications of 2D materials. Though the recently discovered Au-enhanced exfoliation technique provides an effective strategy for preparation of large-scale 2D crystals, the high cost of gold hinders this method from being widely adopted in industrial applications. In addition, direct Au contact could significantly quench photoluminescence (PL) emission in 2D semiconductors. It is therefore crucial to find alternative metals that can replace gold to achieve efficient exfoliation of 2D materials. Here, we present a one-step Ag-assisted method that can efficiently exfoliate many large-area 2D monolayers, where the yield ratio is comparable to Au-enhanced exfoliation method. Differing from Au film, however, the surface roughness of as-prepared Ag films on SiO2/Si substrate is much higher, which facilitates the generation of surface plasmons resulting from the nanostructures formed on the rough Ag surface. More interestingly, the strong coupling between 2D semiconductor crystals (e.g. MoS2, MoSe2) and Ag film leads to a unique PL enhancement that has not been observed in other mechanical exfoliation techniques, which can be mainly attributed to enhanced light-matter interaction as a result of extended propagation of surface plasmonic polariton (SPP). Our work provides a lower-cost and universal Ag-assisted exfoliation method, while at the same offering enhanced SPP-matter interactions.
- Published
- 2022
167. Observer based control for a tree-shaped network of Timoshenko beams using Lyapunov’s method
- Author
-
Meng, Tingting, Zeng, Xiangfu, Wu, Xiaoyang, and Fu, Qiang
- Published
- 2023
- Full Text
- View/download PDF
168. A Strategy for Reducing Nitrogen Fertilizer Application Based on Application of Biochar: A Case in Northeast China Black Soil Region (Mollisols)
- Author
-
Li, Qinglin, Fu, Qiang, Li, Tianxiao, Hou, Renjie, Dong, Shuqi, Xue, Ping, Yang, Xuechen, and Gao, Yu
- Published
- 2023
- Full Text
- View/download PDF
169. The Influence of Meridional Variation in North Pacific Sea Surface Temperature Anomalies on the Arctic Stratospheric Polar Vortex
- Author
-
Wang, Tao, Fu, Qiang, Tian, Wenshou, Liu, Hongwen, Peng, Yifeng, Xie, Fei, Tian, Hongying, and Luo, Jiali
- Published
- 2023
- Full Text
- View/download PDF
170. Dynamics Analysis of the Double Push Rod Limb-Leg Mechanism with Clearance Joint
- Author
-
Zhao, Fu-Qiang, Gao, Zhi-Ying, Chen, Sheng-Qian, Wu, Hong-Qing, Xie, Jia-Quan, Li, Guo-Xing, and Huang, Qing-Xue
- Published
- 2023
- Full Text
- View/download PDF
171. PDK4-dependent hypercatabolism and lactate production of senescent cells promotes cancer malignancy
- Author
-
Dou, Xuefeng, Fu, Qiang, Long, Qilai, Liu, Shuning, Zou, Yejun, Fu, Da, Xu, Qixia, Jiang, Zhirui, Ren, Xiaohui, Zhang, Guilong, Wei, Xiaoling, Li, Qingfeng, Campisi, Judith, Zhao, Yuzheng, and Sun, Yu
- Published
- 2023
- Full Text
- View/download PDF
172. Preoperative Frailty Assessment Predicts Postoperative Mortality, Delirium and Pneumonia in Elderly Lung Cancer Patients: A Retrospective Cohort Study
- Author
-
Tian, Jing-Yang, Hao, Xin-Yu, Cao, Fu-Yang, Liu, Jing-Jing, Li, Yan-Xiang, Guo, Yong-Xin, Mi, Wei-Dong, Tong, Li, and Fu, Qiang
- Published
- 2023
- Full Text
- View/download PDF
173. Development and validation of a predictive model for the diagnosis of bladder tumors using narrow band imaging
- Author
-
Liang, Hao, Yang, Qingya, Zhang, Yaozhong, Sun, Hui, Fu, Qiang, Diao, Tongxiang, Wang, Jin, Huang, Wei, Xu, Yang, Ge, Nan, Jiang, Xuewen, Chen, Shouzhen, Li, Yan, Zhou, Bin, Li, Peixin, Zhang, Xiaoyi, Zhang, Nianzhao, Shi, Benkang, and Chen, Jun
- Published
- 2023
- Full Text
- View/download PDF
174. Study on the Method of Pressure Relief and Energy Absorption for Protecting Roadway Under Thick and Hard Roof
- Author
-
Fu, Qiang, Yang, Jun, Song, Hongxu, Wu, Xing, Liu, Yuxuan, and Wei, Xingjian
- Published
- 2023
- Full Text
- View/download PDF
175. Fine segmentation and difference-aware shape adjustment for category-level 6DoF object pose estimation
- Author
-
Liu, Chongpei, Sun, Wei, Liu, Jian, Zhang, Xing, Fan, Shimeng, and Fu, Qiang
- Published
- 2023
- Full Text
- View/download PDF
176. Concentrated Growth Factor (CGF): The Newest Platelet Concentrate and Its Application in Nasal Hyaluronic Acid Injection Complications
- Author
-
Ding, Hongfan, Fu, Qiang, Liu, Bing, Xv, Xiao, Zhou, Guiwen, Zheng, Can, Chen, Zhaoyang, and Chen, Minliang
- Published
- 2023
- Full Text
- View/download PDF
177. Immortalization-upregulated protein promotes pancreatic cancer progression by regulating NPM1/FHL1-mediated cell-cycle-checkpoint protein activity
- Author
-
Luo, Qiankun, Pan, Yanfeng, Fu, Qiang, Zhang, Xu, Zhou, Shuai, Yu, Pengfei, Tian, Huiyuan, Liu, Pan, Chen, Song, Zhang, Hongwei, and Qin, Tao
- Published
- 2023
- Full Text
- View/download PDF
178. Model Design of Weapon Target Assignment System Based on Multi-Mode Fusion
- Author
-
Liu Xiangyu, Zhu Kun, Wang Gang, Guo Xiangke, Fu Qiang, Li Tengda
- Subjects
weapon target assignment ,us department of defense architecture framework ,petri net ,other allocation ,self-allocation ,command and control ,model design ,Motor vehicles. Aeronautics. Astronautics ,TL1-4050 - Abstract
Target assignment is the core part in the command and control process, and the optimization of the allocation mode is of great significance for improving the air defense and anti-missile combat capability. In order to improve the robustness, applicability, game antagonism and other combat performance of air defense and anti-missile target allocation, so as to cope with the current complex and changeable battlefield environment, this paper proposes the target allocation architecture establishment of multi-mode fusion, and adaptively improve the commercial order-type service mode. The improvement of three kinds of dispatching models, that are “dispatching order” “grabbing order” and “dispatching- grabbing order fusion”, are defined as military models. By using the US department of defense architecture framework (DoDAF) to establish a new target allocation architecture of “other allocation” “self-allocation”and “combination of other allocation and self-allocation”. With the introduction of Perti net model, the reachable map of Petri net model is constructed and analyzed. Through the simulation experiment platform, the complex combat scenario is constructed to verify the feasibility of the distribution strategy mechanism. The results show that each of the three strategies has its own advantages. The multi-strategy combination designed in this paper has greater advantages than the traditional single strategy in matching time, success rate and utility value.
- Published
- 2024
- Full Text
- View/download PDF
179. Dynamic Response of a Vessel-Shaped Fish Cage Considering Coupling Effect Among Body Motion, Disturbed Velocity Field, and Net Loads
- Author
-
WANG Yihou, FU Shixiao, XU Yuwang, LI Shuai, FU Qiang, LIU Fuxiang
- Subjects
vessel-shaped fish cage ,diffraction and radiation waves ,cage motion ,tension in net twine ,volume reduction ,connector load ,Engineering (General). Civil engineering (General) ,TA1-2040 ,Chemical engineering ,TP155-156 ,Naval architecture. Shipbuilding. Marine engineering ,VM1-989 - Abstract
Vessel-shaped fish cages are a new type of large aquaculture structure consisting of a floating body, steel frames, net system, and mooring system. The diffraction and radiation waves induced by the floating body can disturb the velocity field and induce additional changes to the hydrodynamic loads on the nets. In this paper, the velocity transfer functions around the nets induced by the diffraction and radiation waves are obtained and the effects of floating body on the forces of the nets are calculated by the Morison equation. By performing the iterations between the motion of floating body and loads on the nets, the fully coupled dynamic response of motion-disturbing velocity field-net loads is realized. Finally, the effects of diffraction and radiation waves on motion response, tension in the net twine, volume reduction, and connector loads are investigated. The results show that the influences of disturbing velocity field on cage motion response and volume reduction are not obvious, but they can lead to a significant increase in the tension in net twine and connector loads, which can provide helpful reference for the structural strength analysis and safety design of vessel-shaped fish cages.
- Published
- 2024
- Full Text
- View/download PDF
180. Mechanical property of cylindrical sandwich shell with gradient core of entangled wire mesh
- Author
-
Xin Xue, Chao Zheng, Fu-qiang Lai, and Xue-qian Wu
- Subjects
Entangled wire mesh ,Gradient cylindrical sandwich shell ,Vacuum brazing ,Secant stiffness ,Damping ,Military Science - Abstract
To explore the wide-frequency damping and vibration-attenuation performances in the application of aerospace components, the cylindrical sandwich shell structure with a gradient core of entangled wire mesh was proposed in this paper. Firstly, the gradient cores of entangled wire mesh in the axial and radial directions were prepared by using an in-house Numerical Control weaving machine, and the metallurgical connection between skin sheets and the gradient core was performed using vacuum brazing. Secondly, to investigate the mechanical properties of cylindrical sandwich shells with axial or radial gradient cores, quasi-static and dynamic mechanical experiments were carried out. The primary evaluations of mechanical properties include secant stiffness, natural frequency, Specific Energy Absorption (SEA), vibration acceleration level, and so on. The results suggest that the vibration-attenuation performance of the sandwich shell is remarkable when the high-density core layer is at the end of the shell or abuts the inner skin. The axial gradient material has almost no influence on the vibration frequencies of the shell, whereas the vibration frequencies increase dramatically when the high-density core layer approaches the skin. Moreover, compared to the conventional sandwich shells, the proposed functional grading cylindrical sandwich shell exhibits more potential in mass reduction, stiffness designing, and energy dissipation.
- Published
- 2024
- Full Text
- View/download PDF
181. Making Large Language Models Better Reasoners with Step-Aware Verifier
- Author
-
Li, Yifei, Lin, Zeqi, Zhang, Shizhuo, Fu, Qiang, Chen, Bei, Lou, Jian-Guang, and Chen, Weizhu
- Subjects
Computer Science - Computation and Language ,Computer Science - Artificial Intelligence - Abstract
Few-shot learning is a challenging task that requires language models to generalize from limited examples. Large language models like GPT-3 and PaLM have made impressive progress in this area, but they still face difficulties in reasoning tasks such as GSM8K, a benchmark for arithmetic problems. To improve their reasoning skills, previous work has proposed to guide the language model with prompts that elicit a series of reasoning steps before giving the final answer, achieving a significant improvement on GSM8K from 17.9% to 58.1% in problem-solving rate. In this paper, we present DIVERSE (Diverse Verifier on Reasoning Step), a novel approach that further enhances the reasoning capability of language models. DIVERSE has three main components: first, it generates diverse prompts to explore different reasoning paths for the same question; second, it uses a verifier to filter out incorrect answers based on a weighted voting scheme; and third, it verifies each reasoning step individually instead of the whole chain. We evaluate DIVERSE on the latest language model code-davinci-002 and show that it achieves new state-of-the-art results on six of eight reasoning benchmarks (e.g., GSM8K 74.4% to 83.2%).
- Published
- 2022
182. Personalized Acoustic Echo Cancellation for Full-duplex Communications
- Author
-
Zhang, Shimin, Wang, Ziteng, Ju, Yukai, Fu, Yihui, Na, Yueyue, Fu, Qiang, and Xie, Lei
- Subjects
Computer Science - Sound ,Electrical Engineering and Systems Science - Audio and Speech Processing - Abstract
Deep neural networks (DNNs) have shown promising results for acoustic echo cancellation (AEC). But the DNN-based AEC models let through all near-end speakers including the interfering speech. In light of recent studies on personalized speech enhancement, we investigate the feasibility of personalized acoustic echo cancellation (PAEC) in this paper for full-duplex communications, where background noise and interfering speakers may coexist with acoustic echoes. Specifically, we first propose a novel backbone neural network termed as gated temporal convolutional neural network (GTCNN) that outperforms state-of-the-art AEC models in performance. Speaker embeddings like d-vectors are further adopted as auxiliary information to guide the GTCNN to focus on the target speaker. A special case in PAEC is that speech snippets of both parties on the call are enrolled. Experimental results show that auxiliary information from either the near-end speaker or the far-end speaker can improve the DNN-based AEC performance. Nevertheless, there is still much room for improvement in the utilization of the finite-dimensional speaker embeddings., Comment: submitted to INTERSPEECH 22
- Published
- 2022
183. Small Footprint Multi-channel ConvMixer for Keyword Spotting with Centroid Based Awareness
- Author
-
Ng, Dianwen, Pang, Jin Hui, Xiao, Yang, Tian, Biao, Fu, Qiang, and Chng, Eng Siong
- Subjects
Computer Science - Sound ,Electrical Engineering and Systems Science - Audio and Speech Processing - Abstract
It is critical for a keyword spotting model to have a small footprint as it typically runs on-device with low computational resources. However, maintaining the previous SOTA performance with reduced model size is challenging. In addition, a far-field and noisy environment with multiple signals interference aggravates the problem causing the accuracy to degrade significantly. In this paper, we present a multi-channel ConvMixer for speech command recognitions. The novel architecture introduces an additional audio channel mixing for channel audio interaction in a multi-channel audio setting to achieve better noise-robust features with more efficient computation. Besides, we proposed a centroid based awareness component to enhance the system by equipping it with additional spatial geometry information in the latent feature projection space. We evaluate our model using the new MISP challenge 2021 dataset. Our model achieves significant improvement against the official baseline with a 55% gain in the competition score (0.152) on raw microphone array input and a 63% (0.126) boost upon front-end speech enhancement., Comment: submitted to INTERSPEECH 2022
- Published
- 2022
184. Sparse Optical Flow-Based Line Feature Tracking
- Author
-
Fu, Qiang, Yu, Hongshan, Ali, Islam, and Zhang, Hong
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
In this paper we propose a novel sparse optical flow (SOF)-based line feature tracking method for the camera pose estimation problem. This method is inspired by the point-based SOF algorithm and developed based on an observation that two adjacent images in time-varying image sequences satisfy brightness invariant. Based on this observation, we re-define the goal of line feature tracking: track two endpoints of a line feature instead of the entire line based on gray value matching instead of descriptor matching. To achieve this goal, an efficient two endpoint tracking (TET) method is presented: first, describe a given line feature with its two endpoints; next, track the two endpoints based on SOF to obtain two new tracked endpoints by minimizing a pixel-level grayscale residual function; finally, connect the two tracked endpoints to generate a new line feature. The correspondence is established between the given and the new line feature. Compared with current descriptor-based methods, our TET method needs not to compute descriptors and detect line features repeatedly. Naturally, it has an obvious advantage over computation. Experiments in several public benchmark datasets show our method yields highly competitive accuracy with an obvious advantage over speed.
- Published
- 2022
185. Transition edge sensor based detector: from X-ray to $\gamma$-ray
- Author
-
Zhang, Shuo, Xia, Jing-Kai, Sun, Tao, Wu, Wen-Tao, Wu, Bing-Jun, Wang, Yong-Liang, Cantor, Robin, Han, Ke, Zhou, Xiao-Peng, Liu, Hao-Ran, Fan, Fu-You, Guo, Si-Ming, Liang, Jun-Cheng, Li, De-Hong, Song, Yan-Ru, Ju, Xu-Dong, Fu, Qiang, and Liu, Zhi
- Subjects
Physics - Instrumentation and Detectors ,Astrophysics - Instrumentation and Methods for Astrophysics ,81V35 - Abstract
The Transition Edge Sensor is extremely sensitive to the change of temperature, combined with the high-Z metal of a certain thickness, it can realize the high energy resolution measurement of particles such as X-rays. X-rays with energies below 10 keV have very weak penetrating ability, so only a few microns thick of gold or bismuth can obtain quantum efficiency higher than 70\%. Therefore, the entire structure of the TES X-ray detector in this energy range can be realized in the microfabrication process. However, for X-rays or gamma rays from 10 keV to 200 keV, sub-millimeter absorber layers are required, which cannot be realized by microfabrication process. This paper first briefly introduces a set of TES X-ray detectors and their auxiliary systems built by ShanghaiTech University, then focus on the introduction of the TES $\gamma$-ray detector, with absorber based on an sub-millimeter lead-tin alloy sphere. The detector has a quantum efficiency above 70\% near 100 keV, and an energy resolution of about 161.5eV@59.5keV., Comment: 13 pages, 12 figures
- Published
- 2022
186. Nomogram for predicting the overall survival and cancer-specific survival of patients with intraductal carcinoma of the prostate
- Author
-
Cui, Yongqiang, Lin, Junyang, Sun, Dingqi, Zhang, Hui, Diao, Tongxiang, and Fu, Qiang
- Published
- 2024
- Full Text
- View/download PDF
187. Intelligent hydrogel on–off controller sensor for irrigation
- Author
-
Zhan, Xue-Qing, Ran, Zhuo-Qing, Bao, Hong-Yu, Ye, Qin, Chen, Han, Fu, Qiang, Ni, Wang, Xu, Jia-Min, Ma, Ning, and Tsai, Fang-Chang
- Published
- 2024
- Full Text
- View/download PDF
188. A fine-grained reversible data hiding in encrypted domain based on the cipher-text redundancy of encryption process
- Author
-
Yong-jun Kong, Min-qing Zhang, Zong-bao Jiang, Xiong Zhang, Chao Jiang, and Fu-qiang Di
- Subjects
Reversible data hiding in encrypted domain ,Encryption process ,ElGamal encryption ,Fine-grained management ,Blind extraction ,Science (General) ,Q1-390 ,Social sciences (General) ,H1-99 - Abstract
Considering the granularity of embedded data in the design of reversible data hiding scheme has important research significance for the permission control and management of multi-granularity information. To broaden the application possibilities of encrypted data in cloud environments, the researchers propose a fine-grained reversible data hiding method leveraging the cipher-text redundancy of ElGamal encryption. Initially, prior to the encryption process, pixels are organized into a full binary tree based on fine-grained access permissions. Subsequently, a chaotic sequence generator is employed to assign distinct embedding keys to each layer of the full binary tree according to the access permissions. Following this, an XOR operation is conducted between the embedding key and the corresponding secret information in each layer to derive the target features of the cipher-text, facilitating subsequent fine-grained data hiding. Throughout the ElGamal encryption process, iterative manipulation of the random variable ensures alignment between the cipher-text output and the target feature, enabling the embedding of secret information across different layers. This approach facilitates the fine-grained blind extraction of secret information from an encrypted state, thereby expanding the potential applications of cipher-text by extracting information without revealing the original data. Furthermore, the scheme enhances information security through distributed storage and conceals the presence of information hiding by leveraging the separability of lossless decryption and information extraction. Simulation results demonstrate that secret information of three granularities can be embedded and extracted without interference within a three-layer full binary structure, with a maximum embedding capacity of up to 1.75 bpp.
- Published
- 2024
- Full Text
- View/download PDF
189. Three-dimensional printing the navigation template for precise percutaneous renal puncture to treat pyonephrosis on a porcine model and a patient :a case report
- Author
-
Zhang Kaile, Liu Jiafu, Li Wenyao, Yang Xi, Li Ding, Chen Rong, and Fu Qiang
- Subjects
Renal calculi ,Percutaneous nephrolithotomy ,Complications ,Three dimensional printing ,Navigation template ,Science (General) ,Q1-390 ,Social sciences (General) ,H1-99 - Abstract
Objective: Percutaneous nephrolithotomy (PCNL) is the main method for pyonephrosis or lithotripsy in urology. However, it often comes with high risk, as the inaccurate puncture inevitably causes bleeding, intra- and post-operative complications. So, a new inter-disciplinary approach is needed to perform the puncture more accurately. Methods: 3 signs made of lead were marked onto the skin of the posterior side of the waist of a domestic pig or a patient, which was scanned by computed tomography (CT). Based on the CT images, the computer design and the 3D printing, a navigation template made of the transparent resin material is constructed. They were attached onto the surgical area on pig or patient according to the signs. During the PCNL, with this template, the puncture position, angle and depth were optimized in order to precisely enter the targeted renal pelvis or calices. Results: With the 3D navigation templates, 18G puncture needles were used to enter the renal pelvis upon performing the PCNL on a porcine model and a patient. On the porcine model, the urine outflow was observed with minimal complication. Post-operative CT scans revealed that the needle was located in the renal pelvis. For the patient case, the puncture point was designed to target the calix with stone. No obvious bleeding and complication was found in renal puncture with template. Conclusions: The navigation template was made with the combination of 3D printing, CT images and computer design. This template allows for accurate puncture of the renal pelvis or calix. Surgical improvement in kidney stones and pyonephrosis was observed in porcine model and patient case. In the future, prospective, trandomized, controlled clinical trials are needed to further confirm its advantage.
- Published
- 2024
- Full Text
- View/download PDF
190. Pro-inflammatory diets promote the formation of hyperuricemia
- Author
-
Xin Liu, Ting-Yu Chen, Teng-Yu Gao, Ke-Qin Shi, Fu-Qiang Yin, Yun-Xiang Yu, and Chao Zhang
- Subjects
hyperuricemia ,dietary inflammatory index ,NHANES ,drinking ,hypertension ,diabetes ,Diseases of the endocrine glands. Clinical endocrinology ,RC648-665 - Abstract
BackgroundHyperuricemia, as a very prevalent chronic metabolic disease with increasing prevalence year by year, poses a significant burden on individual patients as well as on the global health care and disease burden, and there is growing evidence that it is associated with other underlying diseases such as hypertension and cardiovascular disease. The association between hyperuricemia and dietary inflammatory index (DII) scores was investigated in this study.MethodsThis study enrolled 13, 040 adult subjects (aged ≥ 20 years) from the US National Health and Nutrition Survey from 2003 to 2018. The inflammatory potential of the diet was assessed by the DII score, and logistic regression was performed to evaluate the relationship between the DII score and the development of hyperuricemia; subgroup analyses were used to discuss the influence of other factors on the relationship.ResultsParticipants in the other quartiles had an increased risk of hyperuricemia compared to those in the lowest quartile of DII scores. Stratification analyses stratified by body mass index (BMI), sex, hypertension, drinking, diabetes, education level and albumin-creatinine-ratio (ACR) revealed that the DII score was also associated with the risk of hyperuricemia (P
- Published
- 2024
- Full Text
- View/download PDF
191. A 66‐Nuclear All‐Alkynyl Protected Peanut‐Shaped Silver(I)/Copper(I) Heterometallic Nanocluster: Intermediate in Copper‐Catalyzed Alkyne‐Azide Cycloaddition
- Author
-
Jin‐Ping Gao, Fu‐Qiang Zhang, and Xian‐Ming Zhang
- Subjects
1,3‐dipole cycloadditions ,CuAAC reaction ,Heterometallic nanoclusters ,Intermediate ,Science - Abstract
Abstract Ligand‐protected heterometallic nanoclusters in contrast to homo‐metal counterparts show more broad applications due to the synergistic effect of hetero‐metals but their controllable syntheses remain a challenge. Among heterometallic nanoclusters, monovalent Ag‐Cu compounds are rarely explored due to much difference of Ag(I) and Cu(I) such as atom radius, coordination habits, and redox potential. Encouraged by copper‐catalyzed alkyne‐azide cycloaddition (CuAAC) reaction, comproportionation reaction of Cu(II)X2 and Cu(0) in the presence of (PhC≡CAg)n complex and molybdate generated a core‐shell peanut‐shaped 66‐nuclear Ag(I)‐Cu(I) heterometallic nanocluster, [(Mo4O16)2@Cu12Ag54(PhC≡C)50] (referred to as Ag54Cu12). The structure and composition of Ag‐Cu heterometallic nanocluster are fully characterized. X‐ray single crystal diffraction reveals that Ag54Cu12 has a peanut‐shaped silver(I)/copper(I) heterometallic nanocage protected by fifty phenylacetylene ligands in µ3–modes and encapsulated two mutually twisted tetramolybdates. Heterometallic nanocage contains a 54‐Ag‐atom outer ellipsoid silver cage decorated by 12 copper inside wall. Nanosized Ag54Cu12 is a n‐type narrow‐band‐gap semiconductor with a good photocurrent response. Preliminary experiments demonstrates that Ag54Cu12 itself and activated carbon supported Ag54Cu12/C are effective catalysts for 1,3‐dipole cycloaddition between alkynes and azides at ambient conditions. The work provides not only a new synthetic route toward Ag(I)‐Cu(I) nanoclusters but also an important heterometallic intermediate in CuAAC catalytic reaction.
- Published
- 2024
- Full Text
- View/download PDF
192. Assessment of pesticide exposure to applicators during spraying in orchards with a stretcher-mounted sprayer
- Author
-
Tao Chuanjiang, Mei Chenghan, Zhang Liying, Li Shuang, Yan Yizhou, She Dongmei, An Xuehua, Fu Qiang, Pu Entang, Tao Lingmei, Liu Ran, Zhang Hongjun, and Huang Xiuzhu
- Subjects
Pesticide application ,Occupational exposure ,Orchard spraying ,Stretcher-mounted sprayer ,Risk assessment model ,Science (General) ,Q1-390 ,Social sciences (General) ,H1-99 - Abstract
Various health risk assessment models have been developed to evaluate occupational pesticide exposure in China. However, there has been limited investigation into the relationship between health risks and pesticide spraying in orchards. In this study, we analyzed pesticide exposure of applicators while spraying with a stretcher-mounted sprayer in orchards located in four different climatic regions. All garments’ unit exposure (UE) demonstrated a right-skewed distribution, with gloves and shins accounting for the highest proportion of dermal pesticide exposure. We observed little difference in dermal and inhalation UE levels between apple and citrus orchards, except for pesticide exposure levels on wipes and faces. While 57% of the inhalation UE distribution variance was attributed to clustering and location effects, no significant differences were observed in dermal exposure levels. We evaluated the impact of different levels of protective clothing on pesticide exposure levels, according to applicators' working habits in China. Our findings revealed that improved levels of protection significantly reduced dermal exposure to pesticides, particularly when wearing gloves during spraying with a stretcher-mounted sprayer. Based on our empirical data, we utilized a simple random sampling model and an intercept-only lognormal mixed model to estimate dermal and inhalation exposure levels. The estimated dermal UE was accurate to within 3-fold with 95% confidence, and half of the estimated inhalation UE was acceptable according to the fold relative accuracy (fRA). Our established and verified statistics for dermal and inhalation UE can be utilized to evaluate the potential pesticide exposure to applicators during spraying in orchards with a stretcher-mounted sprayer.
- Published
- 2024
- Full Text
- View/download PDF
193. Two-Dimensional Imaging of Sea-Surface Targets in the Terahertz Band
- Author
-
Li, Xiaofan, primary, Deng, Bin, additional, Fu, Qiang, additional, and Wang, Hongqiang, additional
- Published
- 2023
- Full Text
- View/download PDF
194. Encrypted-SNN: A Privacy-Preserving Method for Converting Artificial Neural Networks to Spiking Neural Networks
- Author
-
Luo, Xiwen, primary, Fu, Qiang, additional, Qin, Sheng, additional, and Wang, Kaiyang, additional
- Published
- 2023
- Full Text
- View/download PDF
195. Input-Tuning: Adapting Unfamiliar Inputs to Frozen Pretrained Models
- Author
-
An, Shengnan, Li, Yifei, Lin, Zeqi, Liu, Qian, Chen, Bei, Fu, Qiang, Chen, Weizhu, Zheng, Nanning, and Lou, Jian-Guang
- Subjects
Computer Science - Computation and Language ,Computer Science - Artificial Intelligence - Abstract
Recently the prompt-tuning paradigm has attracted significant attention. By only tuning continuous prompts with a frozen pre-trained language model (PLM), prompt-tuning takes a step towards deploying a shared frozen PLM to serve numerous downstream tasks. Although prompt-tuning shows good performance on certain natural language understanding (NLU) tasks, its effectiveness on natural language generation (NLG) tasks is still under-explored. In this paper, we argue that one of the factors hindering the development of prompt-tuning on NLG tasks is the unfamiliar inputs (i.e., inputs are linguistically different from the pretraining corpus). For example, our preliminary exploration reveals a large performance gap between prompt-tuning and fine-tuning when unfamiliar inputs occur frequently in NLG tasks. This motivates us to propose input-tuning, which fine-tunes both the continuous prompts and the input representations, leading to a more effective way to adapt unfamiliar inputs to frozen PLMs. Our proposed input-tuning is conceptually simple and empirically powerful. Experimental results on seven NLG tasks demonstrate that input-tuning is significantly and consistently better than prompt-tuning. Furthermore, on three of these tasks, input-tuning can achieve a comparable or even better performance than fine-tuning.
- Published
- 2022
196. HTGN-BTW: Heterogeneous Temporal Graph Network with Bi-Time-Window Training Strategy for Temporal Link Prediction
- Author
-
Yue, Chongjian, Du, Lun, Fu, Qiang, Bi, Wendong, Liu, Hengyu, Gu, Yu, and Yao, Di
- Subjects
Computer Science - Social and Information Networks ,Computer Science - Artificial Intelligence ,Computer Science - Machine Learning - Abstract
With the development of temporal networks such as E-commerce networks and social networks, the issue of temporal link prediction has attracted increasing attention in recent years. The Temporal Link Prediction task of WSDM Cup 2022 expects a single model that can work well on two kinds of temporal graphs simultaneously, which have quite different characteristics and data properties, to predict whether a link of a given type will occur between two given nodes within a given time span. Our team, named as nothing here, regards this task as a link prediction task in heterogeneous temporal networks and proposes a generic model, i.e., Heterogeneous Temporal Graph Network (HTGN), to solve such temporal link prediction task with the unfixed time intervals and the diverse link types. That is, HTGN can adapt to the heterogeneity of links and the prediction with unfixed time intervals within an arbitrary given time period. To train the model, we design a Bi-Time-Window training strategy (BTW) which has two kinds of mini-batches from two kinds of time windows. As a result, for the final test, we achieved an AUC of 0.662482 on dataset A, an AUC of 0.906923 on dataset B, and won 2nd place with an Average T-scores of 0.628942., Comment: 5 pages, Second Winner Award at Temporal Link Prediction task of WSDM Cup 2022
- Published
- 2022
197. MineRL Diamond 2021 Competition: Overview, Results, and Lessons Learned
- Author
-
Kanervisto, Anssi, Milani, Stephanie, Ramanauskas, Karolis, Topin, Nicholay, Lin, Zichuan, Li, Junyou, Shi, Jianing, Ye, Deheng, Fu, Qiang, Yang, Wei, Hong, Weijun, Huang, Zhongyue, Chen, Haicheng, Zeng, Guangjun, Lin, Yue, Micheli, Vincent, Alonso, Eloi, Fleuret, François, Nikulin, Alexander, Belousov, Yury, Svidchenko, Oleg, and Shpilman, Aleksei
- Subjects
Computer Science - Machine Learning ,Computer Science - Artificial Intelligence - Abstract
Reinforcement learning competitions advance the field by providing appropriate scope and support to develop solutions toward a specific problem. To promote the development of more broadly applicable methods, organizers need to enforce the use of general techniques, the use of sample-efficient methods, and the reproducibility of the results. While beneficial for the research community, these restrictions come at a cost -- increased difficulty. If the barrier for entry is too high, many potential participants are demoralized. With this in mind, we hosted the third edition of the MineRL ObtainDiamond competition, MineRL Diamond 2021, with a separate track in which we permitted any solution to promote the participation of newcomers. With this track and more extensive tutorials and support, we saw an increased number of submissions. The participants of this easier track were able to obtain a diamond, and the participants of the harder track progressed the generalizable solutions in the same task., Comment: Under review for PMLR volume on NeurIPS 2021 competitions
- Published
- 2022
198. Multi-Task Deep Residual Echo Suppression with Echo-aware Loss
- Author
-
Zhang, Shimin, Wang, Ziteng, Sun, Jiayao, Fu, Yihui, Tian, Biao, Fu, Qiang, and Xie, Lei
- Subjects
Computer Science - Sound ,Electrical Engineering and Systems Science - Audio and Speech Processing - Abstract
This paper introduces the NWPU Team's entry to the ICASSP 2022 AEC Challenge. We take a hybrid approach that cascades a linear AEC with a neural post-filter. The former is used to deal with the linear echo components while the latter suppresses the residual non-linear echo components. We use gated convolutional F-T-LSTM neural network (GFTNN) as the backbone and shape the post-filter by a multi-task learning (MTL) framework, where a voice activity detection (VAD) module is adopted as an auxiliary task along with echo suppression, with the aim to avoid over suppression that may cause speech distortion. Moreover, we adopt an echo-aware loss function, where the mean square error (MSE) loss can be optimized particularly for every time-frequency bin (TF-bin) according to the signal-to-echo ratio (SER), leading to further suppression on the echo. Extensive ablation study shows that the time delay estimation (TDE) module in neural post-filter leads to better perceptual quality, and an adaptive filter with better convergence will bring consistent performance gain for the post-filter. Besides, we find that using the linear echo as the input of our neural post-filter is a better choice than using the reference signal directly. In the ICASSP 2022 AEC-Challenge, our approach has ranked the 1st place on word accuracy (WAcc) (0.817) and the 3rd place on both mean opinion score (MOS) (4.502) and the final score (0.864)., Comment: ICASSP 2022
- Published
- 2022
- Full Text
- View/download PDF
199. Reasoning Like Program Executors
- Author
-
Pi, Xinyu, Liu, Qian, Chen, Bei, Ziyadi, Morteza, Lin, Zeqi, Fu, Qiang, Gao, Yan, Lou, Jian-Guang, and Chen, Weizhu
- Subjects
Computer Science - Computation and Language ,Computer Science - Artificial Intelligence ,Computer Science - Symbolic Computation - Abstract
Reasoning over natural language is a long-standing goal for the research community. However, studies have shown that existing language models are inadequate in reasoning. To address the issue, we present POET, a novel reasoning pre-training paradigm. Through pre-training language models with programs and their execution results, POET empowers language models to harvest the reasoning knowledge possessed by program executors via a data-driven approach. POET is conceptually simple and can be instantiated by different kinds of program executors. In this paper, we showcase two simple instances POET-Math and POET-Logic, in addition to a complex instance, POET-SQL. Experimental results on six benchmarks demonstrate that POET can significantly boost model performance in natural language reasoning, such as numerical reasoning, logical reasoning, and multi-hop reasoning. POET opens a new gate on reasoning-enhancement pre-training, and we hope our analysis would shed light on the future research of reasoning like program executors., Comment: To appear in EMNLP 2022 main conference. The first two authors contributed equally
- Published
- 2022
200. ConvMixer: Feature Interactive Convolution with Curriculum Learning for Small Footprint and Noisy Far-field Keyword Spotting
- Author
-
Ng, Dianwen, Chen, Yunqi, Tian, Biao, Fu, Qiang, and Chng, Eng Siong
- Subjects
Computer Science - Sound ,Electrical Engineering and Systems Science - Audio and Speech Processing - Abstract
Building efficient architecture in neural speech processing is paramount to success in keyword spotting deployment. However, it is very challenging for lightweight models to achieve noise robustness with concise neural operations. In a real-world application, the user environment is typically noisy and may also contain reverberations. We proposed a novel feature interactive convolutional model with merely 100K parameters to tackle this under the noisy far-field condition. The interactive unit is proposed in place of the attention module that promotes the flow of information with more efficient computations. Moreover, curriculum-based multi-condition training is adopted to attain better noise robustness. Our model achieves 98.2% top-1 accuracy on Google Speech Command V2-12 and is competitive against large transformer models under the designed noise condition., Comment: submitted to ICASSP 2022
- Published
- 2022
- Full Text
- View/download PDF
Catalog
Discovery Service for Jio Institute Digital Library
For full access to our library's resources, please sign in.