检索结果-内蒙古大学图书馆

Multilingual Simultaneous Speech Translation

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Subramanya, Shashank Niehues, Jan Department of Data Science and Knowledge Engineering Maastricht University Netherlands Institute for Anthropomatics and Robotics Karlsruhe Institute of Technology Germany

Applications designed for simultaneous speech translation during events such as conferences or meetings need to balance quality and lag while displaying translated text to deliver a good user experience. One common approach to building online spoken language translation systems is by leveraging models built for offline speech translation. Based on a technique to adapt end-to-end monolingual models, we investigate multilingual models and different architectures (end-to-end and cascade) on the ability to perform online speech translation. On the multilingual TEDx corpus, we show that the approach generalizes to different architectures. We see similar gains in latency reduction (40% relative) across languages and architectures. However, the end-to-end architecture leads to smaller translation quality losses after adapting to the online model. Furthermore, the approach even scales to zero-shot directions. © 2022, CC BY.

关键词： Online systems

Cyclic generators and an improved linear kernel for the rooted subtree prune and regraft distance

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Kelk, Steven Linz, Simone Meuwese, Ruben Department of Data Science and Knowledge Engineering Maastricht University Netherlands School of Computer Science University of Auckland New Zealand

The rooted subtree prune and regraft (rSPR) distance between two rooted binary phylogenetic trees is a well-studied measure of topological dissimilarity that is NP-hard to compute. Here we describe an improved linear kernel for the problem. In particular, we show that if the classical subtree and chain reduction rules are augmented with a modified type of chain reduction rule, the resulting trees have at most 9k−3 leaves, where k is the rSPR distance;and that this bound is tight. The previous best-known linear kernel had size O(28k). To achieve this improvement we introduce cyclic generators, which can be viewed as cyclic analogues of the generators used in the phylogenetic networks literature. As a corollary to our main result we also give an improved weighted linear kernel for the minimum hybridization problem on two rooted binary phylogenetic trees. Copyright © 2022, The Authors. All rights reserved.

关键词： Bioinformatics

Efficient Speech Translation with Pre-trained Models

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Li, Zhaolin Niehues, Jan Department of Data Science and Knowledge Engineering Maastricht University Netherlands Institute for Anthropomatics and Robotics Karlsruhe Institute of Technology Germany

When building state-of-the-art speech translation models, the need for large computational resources is a significant obstacle due to the large training data size and complex models. The availability of pre-trained models is a promising opportunity to build strong speech translation systems efficiently. In a first step, we investigate efficient strategies to build cascaded and end-to-end speech translation systems based on pre-trained models. Using this strategy, we can train and apply the models on a single GPU. While the end-to-end models show superior translation performance to cascaded ones, the application of this technology has a limitation on the need for additional end-to-end training data. In a second step, we proposed an additional similarity loss to encourage the model to generate similar hidden representations for speech and transcript. Using this technique, we can increase the data efficiency and improve the translation quality by 6 BLEU points in scenarios with limited end-to-end training data. © 2022, CC BY.

关键词： Speech

Collaborative Hybrid Human AI Learning through Conceptual Exploration 3

学校读者我要写书评

暂无评论

Collaborative Hybrid Human AI Learning through Conceptual Ex...

3rd International Conference on Hybrid Human-Artificial Intelligence, HHAI-WS 2024

作者： Ganter, Bernhard Hanika, Tom Hirth, Johannes Obiedkov, Sergei Ernst-Schröder-Zentrum Karolinenpl. 5 Darmstadt64289 Germany TU Dresden Dresden01062 Germany Intelligent Information Systems Universisty of Hildesheim Universitätsplatz 1 Hildesheim31141 Germany Knowledge & Data Engineering Group University of Kassel Wilhelmshöher Allee 73 Kassel34121 Germany Faculty of Computer Science cfaed ScaDS.AI TU Dresden Dresden01062 Germany

Conceptual Exploration is a sophisticated method for the interactive and structured acquisition of knowledge from experts. It is therefore particularly suitable for the use in hybrid settings where both humans and AIs act as experts. This article provides a brief summary of how conceptual exploration can be used in the context of Hybrid Human AI systems, as discussed within a tutorial during the third HHAI conference in Malmö, Sweden. We will recapitulate two small experiments that were carried out with the participants of the tutorial and their results. Finally, we give some pointers on how this promising link can be further researched. © 2024 Copyright for this paper by its authors.

关键词： Contrastive Learning

Task Offloading Optimization Based on Position Prediction in Pedestrian- Robot Mixed Traffic Flows

学校读者我要写书评

暂无评论

Task Offloading Optimization Based on Position Prediction in...

International Conference on Computer and Communications (ICCC)

作者： Dawen Zheng Lusheng Wang Caihong Kai Min Peng Key Laboratory of Knowledge Engineering with Big Data Ministry of Education School of Computer Science and Information Engineering Hefei University of Technology Hefei China

Service robots play an increasingly important role in people's daily life. The density of pedestrians is large and the movement is irregular in pedestrian-robot mixed traffic flows. Robots are prone to collision with pedestrians, and the tasks to be offloaded are closely related to pedestrians. How to analyze the tasks of robots and select the appropriate roadside unit is an important issue. In this paper, the social force model is used to predict the positions of pedestrians and robots, taking into account the influence of various forces to avoid collisions. A task offloading resource optimization algorithm with position prediction is proposed. According to the predicted information, the size and position distribution of all tasks in the scenario are obtained, and then the neural network trained beforehand based on deep Q-Iearning is used to generate a task offloading strategy. The simulation results show that the running time of the proposed algorithm is very short, and the resource allocation required for task offloading is completed in advance based on the predicted information before robots arriving the corresponding positions. Besides, the algorithm significantly reduces the task offloading delay.

关键词： Service robots Simulation Force Predictive models Prediction algorithms Delays Task analysis

An accurate diagnostic model for PNS in children based on random wandering and vector space Constructing research*

学校读者我要写书评

暂无评论

An accurate diagnostic model for PNS in children based on ra...

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Kui Zhang ShuangGui Tian Yong Xiao ShaoWu Shen ShuSong Mao College of Information Engineering Hubei University of Chinese Medicine Wuhan China Research Institute for Traditional Chinese Medicine Knowledge Engineering and Big Data Hubei University of Chinese Medicine Wuhan China Hubei Provincial Hospital of Traditional Chinese Medicine Wuhan China

ISBN: (数字)9798350386226

ISBN: (纸本)9798350386233

Pediatric nephrotic syndrome (PNS) is a common urological disease in children, and one of the main symptoms of PNS is spleen and kidney yang deficiency. The aim of this study was to construct a 5-layer accurate diagnostic model, adding disease staging and syndromic grouping to the three diagnostic layers of "disease, evidence and symptom". First, we use the random walk method for the topological network of primary symptoms and the feature group similarity calculation method for secondary symptoms to match the optimal staging of cases. Then, we use the cosine similarity between the vector space of each evidence grouping and the case ensemble vector to select the optimal evidence grouping. This approach provides new ideas and methods for clinical individualized treatment in traditional Chinese medicine (TCM).

关键词： Spleen knowledge engineering Pediatrics Big data Vectors data models Clinical diagnosis Medical diagnostic imaging Kidney Diseases

Uncovering the Impact of Chain-of-Thought Reasoning for Direct Preference Optimization: Lessons from Text-to-SQL

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Liu, Hanbing Li, Haoyang Zhang, Xiaokang Chen, Ruotong Xu, Haiyong Tian, Tian Qi, Qi Zhang, Jing Gaoling School of Artificial Intelligence Renmin University of China Beijing China School of Information Renmin University of China Beijing China Key Laboratory of Data Engineering and Knowledge Engineering Beijing China Engineering Research Center of Database and Business Intelligence Beijing China China Mobile Information Technology Center China

Direct Preference Optimization (DPO) has proven effective in complex reasoning tasks like math word problems and code generation. However, when applied to Text-to-SQL datasets, it often fails to improve performance and can even degrade it. Our investigation reveals the root cause: unlike math and code tasks, which naturally integrate Chain-of-Thought (CoT) reasoning with DPO, Text-to-SQL datasets typically include only final answers (gold SQL queries) without detailed CoT solutions. By augmenting Text-to-SQL datasets with synthetic CoT solutions, we achieve, for the first time, consistent and significant performance improvements using DPO. Our analysis shows that CoT reasoning is crucial for unlocking DPO’s potential, as it mitigates reward hacking, strengthens discriminative capabilities, and improves scalability. These findings offer valuable insights for building more robust Text-to-SQL models. To support further research, we publicly release the code and CoT-enhanced datasets © 2025, CC BY.

关键词： Chains

UAV-assisted Joint Mobile Edge Computing and data Collection via Matching-enabled Deep Reinforcement Learning

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Wang, Boxiong Kang, Hui Li, Jiahui Sun, Geng Sun, Zemin Wang, Jiacheng Niyato, Dusit College of Computer Science and Technology Jilin University Changchun130012 China Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin University Changchun130012 China College of Computing and Data Science Nanyang Technological University 639798 Singapore College of Computing and Data Science Nanyang Technological University Singapore

Unmanned aerial vehicle (UAV)-assisted mobile edge computing (MEC) and data collection (DC) have been popular research issues. Different from existing works that consider MEC and DC scenarios separately, this paper investigates a multi-UAV-assisted joint MEC-DC system. Specifically, we formulate a joint optimization problem to minimize the MEC latency and maximize the collected data volume. This problem can be classified as a non-convex mixed integer programming problem that exhibits long-term optimization and dynamics. Thus, we propose a deep reinforcement learning-based approach that jointly optimizes the UAV movement, user transmit power, and user association in real time to solve the problem efficiently. Specifically, we reformulate the optimization problem into an action space-reduced Markov decision process (MDP) and optimize the user association by using a two-phase matching-based association (TMA) strategy. Subsequently, we propose a soft actor-critic (SAC)-based approach that integrates the proposed TMA strategy (SAC-TMA) to solve the formulated joint optimization problem collaboratively. Simulation results demonstrate that the proposed SAC-TMA is able to coordinate the two subsystems and can effectively reduce the system latency and improve the data collection volume compared with other benchmark algorithms. Copyright © 2025, The Authors. All rights reserved.

关键词： Unmanned aerial vehicles (UAV)

Enhancing Playout Policy Adaptation for General Game Playing 1

学校读者我要写书评

暂无评论

1st Workshop on Monte Carlo Search, MCS 2020, Held in conjunction with IJCAI 2020

作者： Sironi, Chiara F. Cazenave, Tristan Winands, Mark H. M. Game AI & Search Group Department of Data Science and Knowledge Engineering Maastricht University Maastricht Netherlands LAMSADE Université Paris-Dauphine PSL CNRS Paris France

ISBN: (数字)9783030894535

ISBN: (纸本)9783030894528

Playout Policy Adaptation (PPA) is a state-of-the-art strategy that has been proposed to control the playouts in Monte-Carlo Tree Search (MCTS). PPA has been successfully applied to many two-player, sequential-move games. This paper further evaluates this strategy in General Game Playing (GGP) by first reformulating it for simultaneous-move games. Next, it presents five enhancements for the strategy, four of which have been previously successfully applied to a related MCTS playout strategy, the Move-Average Sampling Technique (MAST). Experiments on a heterogeneous set of games show three enhancements to have a positive effect on PPA: (i) updating the policy for all players proportionally to their payoffs instead of updating only the policy of the winner, (ii) collecting statistics for N-grams of moves instead of single moves only, and (iii) discounting the backpropagated payoffs depending on the depth of the playout. Results also show enhanced PPA variants to be competitive with MAST for small search budgets, and better for larger search budgets. The use of an ϵ -greedy selection of moves and of after-move decay of statistics, instead, seem to have a detrimental effect on PPA. © 2021, Springer Nature Switzerland AG.

关键词： Monte Carlo methods