Article
Computer Science, Artificial Intelligence
Tonnes F. Nygaard, Charles P. Martin, Jim Torresen, Kyrre Glette, David Howard
Summary: The study introduces the first quadrupedal robot capable of morphologically adapting to different environmental conditions in outdoor, unstructured environments. Through embodied AI and an adaptation algorithm, the robot transitions between the most energy-efficient morphologies based on currently sensed terrains, showing significant performance improvements over non-adaptive approaches. This demonstration highlights the potential for a new embodied way of incorporating adaptation into future robotic designs.
NATURE MACHINE INTELLIGENCE
(2021)
Article
Computer Science, Information Systems
Hua Li, Jianmei Duan, Qiubai Sun
Summary: This study improves the traditional public goods game model by combining game theory and interactive learning, and proposes a strategy adaptive evolution method based on a public goods game. The results show that moderate weight adjustment can effectively facilitate cooperative evolution, and the level of cooperation depends mainly on the weight distribution of participants.
Article
Multidisciplinary Sciences
Natalia Kowalczyk-Grebska, Maciek Skorko, Pawel Dobrowolski, Bartosz Kossowski, Monika Mysliwiec, Nikodem Hryniewicz, Maciej Gaca, Artur Marchewka, Malgorzata Kossut, Aneta Brzezicka
Summary: This research found that players' brain structures may be related to real-time strategy game skills, aiding in successful learning of new skills, not limited to video games, but can also be applied to broader cognitive training.
ANNALS OF THE NEW YORK ACADEMY OF SCIENCES
(2021)
Article
Multidisciplinary Sciences
Janusz Kloskowski
Summary: Understanding animal responses to environmental change is crucial for management of ecological traps. The study found that red-necked grebes were more likely to switch territories to ponds with high habitat quality from the previous year, regardless of the current quality. However, due to rotation of fish stocks, many ponds experienced changes in habitat quality the following year, trapping birds who had made decisions based on past information.
SCIENTIFIC REPORTS
(2021)
Article
Computer Science, Cybernetics
Abdessamed Ouessai, Mohammed Salem, Antonio M. Mora
Summary: Real-Time Strategy games pose challenges for search and machine learning due to their large combinatorial decision and state spaces. Exploiting domain knowledge can assist in navigating these spaces and improve game-playing agents' performance.
ENTERTAINMENT COMPUTING
(2022)
Article
Computer Science, Artificial Intelligence
Gia Thuan Lam, Doina Logofatu, Costin Badica
Summary: Real-time fighting games pose a challenge for computer agents due to the necessity of quick decision-making, achieved by powerful machines or advanced algorithms. This paper focuses on algorithmic approaches in real-time fighting games, proposing generic heuristics in combination with Monte-Carlo tree search that outperforms traditional algorithms. The research suggests that the proposed heuristics may have applications beyond the scope of fighting game AI challenges.
Article
Chemistry, Analytical
Hong-Wei Li, Fang Wang, Yi-Qing Ni, You-Wu Wang, Zhao-Dong Xu
Summary: This paper proposes a novel RTHS control strategy that combines the theories of adaptive control and robust control to simplify the control system design and achieve the adaptability and robustness of the system.
Review
Radiology, Nuclear Medicine & Medical Imaging
Margarita Kirienko, Martina Sollini, Gaia Ninatti, Daniele Loiacono, Edoardo Giacomello, Noemi Gozzi, Francesco Amigoni, Luca Mainardi, Pier Luca Lanzi, Arturo Chiti
Summary: This scoping review aimed to assess the non-inferiority of distributed learning compared to centrally and locally trained machine learning models in medical applications. The study found that distributed learning performed close to centralized training and outperformed locally trained models in most cases, indicating its potential significance for ML-based research and practice.
EUROPEAN JOURNAL OF NUCLEAR MEDICINE AND MOLECULAR IMAGING
(2021)
Article
Computer Science, Information Systems
S. Muruganandam, J. Arokia Renjit
Summary: This paper presents an efficient Real-time Reliable Clustering and Secure Transmission (RRCST) approach for Mobile Adhoc Networks (MANET), improving data security and QoS performance through clustering nodes and selecting routes based on various support measures.
PEER-TO-PEER NETWORKING AND APPLICATIONS
(2021)
Article
Agriculture, Multidisciplinary
Tianhai Wang, Ning Wang, Jianxing Xiao, Yanlong Miao, Yifan Sun, Han Li, Man Zhang
Summary: This paper proposes a novel one-shot domain adaptive real-time 3D obstacle detection method based on semantic-geometry-intensity fusion strategy. By introducing the concept of one-shot domain adaptation, the proposed method enables fine-grained 3D obstacle detection with just one sample per category.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2023)
Article
Computer Science, Information Systems
Anshul Thakur, Jacob Armstrong, Alexey Youssef, David Eyre, David A. Clifton
Summary: Healthcare is a dynamic field, and clinical AI models often become ineffective due to the evolving demographics, diseases, and therapeutics. Incremental learning is an effective method to adapt these models to distribution shifts, but it can be unreliable as any adverse modification can render the model unsuitable. This paper introduces self-aware SGD, an incremental deep learning algorithm that utilizes a contextual bandit-like sanity check to ensure reliable modifications to a model.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Article
Computer Science, Artificial Intelligence
Hsuan-Min Wang, Chia-Yuan Hou, Chuen-Tsai Sun
Summary: This article discusses the characteristics of real-time strategy (RTS) games and presents a study on game design and player application of strategies and tactics.
IEEE TRANSACTIONS ON GAMES
(2022)
Article
Computer Science, Information Systems
Minghao Wang, Yongjie Ma, Peidi Wang
Summary: This paper proposes a parameter and strategy adaptive differential evolution algorithm based on accompanying evolution (APSDE). By optimizing the accompanying population, the strategy and parameters of the main population are adapted, and population diversity is enhanced by generating reverse individuals.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Jose Aguilar, Francisco Diaz, Junior Altamiranda, Nelson Perez Garcia, Angel Dario Pinto
Summary: This article introduces a new component of an emerging serious game engine (ESGE), called the plot adaptive system (PAS), which allows updating game plots according to the given theme in a smart classroom. The article also analyzes the behavior of PAS in a case study and presents three experiments with very encouraging results.
IEEE TRANSACTIONS ON GAMES
(2022)
Article
Robotics
Nicolas Harvey Chapman, Feras Dayoub, Will Browne, Christopher Lehnert
Summary: Unsupervised Domain Adaptive Object Detection (UDA-OD) uses unlabelled data to improve the reliability of robotic vision systems in open-world environments. We propose a framework that explicitly addresses class distribution shift to improve pseudo-label reliability in self-training. Our method utilizes the domain invariance and contextual understanding of a pre-trained joint vision and language model to predict the class distribution of unlabelled data, and adjusts the pseudo-labels based on this prediction, providing weak supervision of pseudo-label accuracy. Additionally, we dynamically adjust the number of pseudo-labels per image based on model confidence to account for low quality pseudo-labels early in self-training. Our method outperforms state-of-the-art approaches on multiple benchmarks, including a 4.7 mAP improvement when facing challenging class distribution shift.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2023)