Article
Computer Science, Artificial Intelligence
Gyoung S. Na
Summary: This paper proposes a new approach to improve the optimization performance of gradient-based optimizers in metric learning by using eigenguidance, which is calculated based on eigenvalue decomposition. The experiments showed that the gradient-based optimizers with eigenguidance converged significantly faster than those without in metric learning tasks.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Chemistry, Multidisciplinary
Yupeng Li, Huimin Lu, Yifan Wang, Ruoran Gao, Chengcheng Zhao
Summary: In this study, a new model combining the vision transformer architecture with the capsule network (ViT-Cap) was proposed for finger vein recognition. The model explores finger vein image information based on global and local attention and selectively focuses on important finger vein feature information. Experimental results showed that the proposed model achieved better recognition accuracy compared to the original vision transformer, capsule network, and other advanced finger vein recognition algorithms. Moreover, the model achieved state-of-the-art performance in terms of equal error rate (EER), particularly on the FV-USM datasets, demonstrating its effectiveness and reliability in finger vein recognition.
APPLIED SCIENCES-BASEL
(2022)
Review
Computer Science, Artificial Intelligence
Anubha Parashar, Apoorva Parashar, Weiping Ding, Rajveer S. S. Shekhawat, Imad Rida
Summary: This paper provides a comprehensive overview of deep learning architectures and pipelines for biometric applications using complex characteristics of human gait. The authors discuss the challenges in gait recognition due to various covariates and present a literature review on the performance of deep learning models in covariate conditions. They also cover various aspects of deep learning pipelines in gait recognition, such as data acquisition, preprocessing, feature extraction, and classification. The paper concludes by highlighting the benefits and drawbacks of deep learning approaches in covariate conditions and identifying open problems in identification based on behavioral traits.
ARTIFICIAL INTELLIGENCE REVIEW
(2023)
Review
Veterinary Sciences
Sigfredo Fuentes, Claudia Gonzalez Viejo, Eden Tongson, Frank R. Dunshea
Summary: Livestock welfare assessment plays a crucial role in monitoring animal health, maintaining productivity, and responding to consumer demand for humane treatment. Recent advancements in remote sensing, computer vision, and AI have enabled the development of new technologies for extracting key physiological parameters associated with animal welfare. However, there is a need for more practical applications and validation of these methods, as well as the development of efficient and non-contact AI-based approaches.
ANIMAL HEALTH RESEARCH REVIEWS
(2022)
Article
Computer Science, Information Systems
Zeu Kim, Youngin Kim, Young-Joo Suh
Summary: Development of deep learning has improved computer vision tasks such as image retrieval. This paper introduces a descriptor mixer that combines local and global descriptors for enhanced performance. The combination of GeM and MAC achieved the best results, improving the model's performance by 1.36% (recall @ 32).
Article
Computer Science, Artificial Intelligence
Cong Liu, Wenhao She, Minjie Chen, Xiaofang Li, Simon X. Yang
Summary: Zero-shot image retrieval is the task of retrieving images of unseen classes using a query image of the same class. Existing methods for zero-shot image retrieval focus on pushing the decision boundary between intra-class and inter-class similarities. However, using a universal threshold in the inference stage can compromise performance. To address this, we propose a novel Consistent Penalizing Field (CPF) Loss that creates consistent decision boundaries for all classes. Experimental results show that the proposed method outperforms state-of-the-art methods on various datasets.
EXPERT SYSTEMS WITH APPLICATIONS
(2024)
Review
Agriculture, Multidisciplinary
Yuzhen Lu, Dong Chen, Ebenezer Olaniyi, Yanbo Huang
Summary: This paper presents an overview of the application of generative adversarial networks (GANs) in agricultural image augmentation or synthesis to improve model performance. GANs have been applied in various visual recognition tasks in agriculture and food systems, including plant health conditions, weeds, fruits (preharvest), aquaculture, animal farming, plant phenotyping, and postharvest detection of fruit defects, demonstrating numerous applications and potential opportunities.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2022)
Article
Construction & Building Technology
Hyunjun Kim, Sung-Han Sim, Billie F. Spencer
Summary: This study introduces an advanced stereo vision framework using wide-angle and telephoto lenses for crack quantification and 3D reconstruction of concrete structures. A robust depth estimation strategy is also proposed and the performance was field validated.
AUTOMATION IN CONSTRUCTION
(2022)
Article
Agronomy
Habib Khan, Ijaz Ul Haq, Muhammad Munsif, Mustaqeem, Shafi Ullah Khan, Mi Young Lee
Summary: Agriculture is an important sector of human life, and wheat, as the most farmed crop, is often affected by diseases. To improve disease recognition and increase yield, researchers have proposed an efficient machine learning-based framework, which includes data collection, preprocessing, and model training.
Article
Surgery
Thomas M. Ward, Daniel A. Hashimoto, Yutong Ban, Guy Rosman, Ozanan R. Meireles
Summary: The study found that an AI model can accurately identify the degree of gallbladder inflammation, which has an impact on the intra-operative course. The automated assessment system can be used for optimizing the workflow in the operating room and providing targeted feedback to surgeons and residents, accelerating the acquisition of operative skills.
SURGICAL ENDOSCOPY AND OTHER INTERVENTIONAL TECHNIQUES
(2022)
Review
Agriculture, Multidisciplinary
Yunchao Tang, Jiajun Qiu, Yunqi Zhang, Dongxiao Wu, Yuhong Cao, Kexin Zhao, Lixue Zhu
Summary: The demand for intelligent agriculture is increasing due to global food and environmental crises. This article reviews optimization strategies used in fruit detection and explores methods for improving fruit detection in complex environments. By focusing on the challenges posed by unstructured orchard environments, the effectiveness of optimization measures before and after image sampling are compared. The future development trends of fruit detection optimization technology in complex backgrounds are also discussed.
PRECISION AGRICULTURE
(2023)
Article
Computer Science, Artificial Intelligence
Quan Cui, Zhao-Min Chen, Osamu Yoshie
Summary: This paper investigates the representation learning problem of deep hashing in the nearest neighbor search. Experimental results demonstrate that although deep hashing can accelerate query speed and reduce storage cost, it sacrifices the discriminability of deep representations. To address this problem, a two-step deep hashing learning framework is proposed, which can simultaneously learn compact binary codes and protect deep representations from being sacrificed.
Article
Robotics
Claus Smitt, Michael Halstead, Alireza Ahmadi, Chris McCool
Summary: In agriculture, most vision systems focus on still image classification. However, recent research has shown that spatial and temporal cues have the potential to enhance classification performance. This letter introduces novel approaches that explicitly capture spatial and temporal information to improve the classification of deep convolutional neural networks. By utilizing RGB-D images and robot odometry, the inter-frame feature map spatial registration is performed, and this information is integrated into recurrent deep learned models to enhance their accuracy and robustness. The results demonstrate a substantial improvement in classification performance using our best spatial-temporal model.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2022)
Article
Plant Sciences
Rong Tang, Yujie Lei, Beisiqi Luo, Junbo Zhang, Jiong Mu
Summary: This paper proposes an efficient plum fruit detection model based on an improved YOLOv7 algorithm. The model can quickly and accurately detect plum fruits in complex orchard environments. By capturing high-resolution images of plum fruits growing in natural conditions and forming a dataset through manual screening, data enhancement, and annotation, the YOLOv7-plum algorithm achieved an average precision (AP) value of 94.91%, a 2.03% improvement compared to the YOLOv7 model. The experimental results demonstrate the better performance of the proposed method in detecting plum fruits in complex backgrounds, which contributes to the development of intelligent cultivation in the plum industry.
Article
Plant Sciences
Jia Yao, Yubo Wang, Ying Xiang, Jia Yang, Yuhang Zhu, Xin Li, Shuangshuang Li, Jie Zhang, Guoshu Gong
Summary: Prevention and management of crop diseases are crucial in agricultural production. This paper focuses on common diseases of kiwifruit and proposes an innovative method using deep learning and computer vision models to accurately identify diseases. Experimental results demonstrate high accuracy and robustness of the proposed method.
Article
Agriculture, Multidisciplinary
Siobhan Mullan, Philippa Wiltshire, Kate Cross, David C. J. Main, Kate Still, Madeleine Crawley, Andrew W. Dowsey
Summary: Animal welfare is a crucial aspect of sustainability in livestock farming, and the introduction of welfare outcome monitoring in the Red Tractor UK national dairy assurance scheme has led to significant improvements in reducing the prevalence of diseases among cows while increasing milk yield per cow.
INTERNATIONAL JOURNAL OF AGRICULTURAL SUSTAINABILITY
(2022)
Article
Veterinary Sciences
A. Quain, S. Mullan, M. P. Ward
Summary: The COVID-19 pandemic has led to significant changes in veterinary practice communication, with challenges such as reduced face-to-face contact, difficulties communicating while wearing personal protective equipment, and convincing clients of new protocols. Veterinary teams need to modify communication strategies to facilitate effective communication in the current situation.
AUSTRALIAN VETERINARY JOURNAL
(2022)
Article
Agriculture, Dairy & Animal Science
Elizabeth Rowe, Siobhan Mullan
Summary: A good life for farmed animals is achieved by providing valued resources and positive physical and mental experiences. Evaluating resource provision is crucial for assessing positive welfare for farm animals.
Article
Agriculture, Dairy & Animal Science
Anne Quain, Siobhan Mullan, Michael P. Ward
Summary: During the COVID-19 pandemic, low and no-contact euthanasia have posed common and/or stressful ethical challenges in veterinary practices. A toolkit of protocols is recommended to assist veterinary team members in performing low-contact euthanasia and avoiding no-contact euthanasia whenever possible, to minimize negative impacts on team members, clients, and animal patients.
Article
Veterinary Sciences
A. S. Cooke, S. M. Mullan, C. Morten, J. Hockenhull, M. R. F. Lee, L. M. Cardenas, M. J. Rivero
Summary: Animal welfare is an integral part of livestock production and sustainability. Qualitative Behaviour Assessment (QBA) is a technique used to assess animal welfare, with behavior being a key component. Video-QBA (V-QBA) is a method that uses video footage for assessment, and studies have found broad agreement between V-QBA and live QBA results. However, caution should be taken when implementing V-QBA due to the lack of absolute agreement and lower scores obtained in some cases.
FRONTIERS IN VETERINARY SCIENCE
(2022)
Article
Veterinary Sciences
Anne Quain, Siobhan Mullan, Michael P. Ward
Summary: This study evaluated the impact of ethics rounds, a form of clinical ethics support service (CESS), on veterinary team members. The results showed that participating in ethics rounds can improve the ability of veterinary team members to recognize and navigate ethically challenging situations (ECS), potentially mitigating moral distress.
FRONTIERS IN VETERINARY SCIENCE
(2022)
Article
Agriculture, Dairy & Animal Science
Helena Hale, Emily Blackwell, Claire Roberts, Emma Roe, Siobhan Mullan
Summary: Despite the availability of formal tools for veterinary assessment of canine quality of life, they are rarely used in practice due to perceived resistance from dog owners. However, an online survey suggests that the majority of UK dog owners are comfortable discussing their dogs' quality of life with their vets and are interested in accessing assessment tools. Interviews with a subset of owners further confirm their desire to have holistic dog care discussions and to use formal assessment tools. These findings suggest that the use of tools can improve the vet-client relationship and owner confidence in dog treatment.
Article
Agriculture, Dairy & Animal Science
Nicola J. Rooney, Paula E. Baker, Emily-Jayne Blackwell, Matthew G. Walker, Siobhan Mullan, Ricahrd A. Saunders, Suzanne D. E. Held
Summary: Although studies on the housing needs of laboratory and meat rabbits are available, there is a lack of research on the requirements of pet rabbits, particularly those kept in pairs. This study found that small hutches and restricted access to an exercise area can lead to increased stress hormone levels in pet rabbits, highlighting the importance of providing rabbits with sufficient exercise freedom.
APPLIED ANIMAL BEHAVIOUR SCIENCE
(2023)
Review
Veterinary Sciences
Kate Allen, Lynley Anderson, Mike King, Siobhan Mullan
Summary: This scoping review examines the existing literature on equine sports medicine ethics to identify current concerns and issues, and to map areas for future research. The review finds that the literature mainly focuses on competing stakeholder interests, governing bodies and regulations, provision of optimal veterinary care, confidentiality, and social license for the veterinary profession. The review calls for further consideration on how the veterinary profession and sporting governing bodies can support veterinary surgeons to strive for the highest levels of professional conduct and establish processes for determining ethical veterinary practices.
EQUINE VETERINARY JOURNAL
(2023)
Article
Veterinary Sciences
I Chan, A. Dowsey, P. Lait, S. Tasker, E. Blackwell, C. R. Helps, E. N. Barker
Summary: This study investigates the common causes of upper respiratory tract disease (URTD) in cats in the UK pet cat population and the risk factors for their oral carriage. The results showed that out of 430 cats, 9 (2.1%) were positive for feline herpesvirus (FHV), 57 (13.3%) were positive for feline calicivirus (FCV), and 5 (1.2%) were positive for Chlamydia felis. FCV was the most frequently encountered URTD pathogen in this sample of cats, highlighting the importance of appropriate disinfectant choice. Assessment for co-infection with FCV is recommended in cats suspected of having FHV or C. felis infection.
JOURNAL OF SMALL ANIMAL PRACTICE
(2023)
Article
Multidisciplinary Sciences
Leah E. Trigg, Sally Lyons, Siobhan Mullan
Summary: This study identified risk factors associated with the occurrence of exertional heat illness (EHI) in racehorses, including race distance, wet bulb globe temperature, preceding 5-day temperature average, occurrence of a previous EHI incident, going, year, and race off time. The results provide important evidence for the industry to implement measures such as providing appropriate cool down facilities, early intervention for horses with repeated EHI incidents, and collecting new data streams like on-course wet bulb globe temperature measurements.
SCIENTIFIC REPORTS
(2023)
Article
Agriculture, Dairy & Animal Science
Claire Roberts, Emily J. Blackwell, Emma Roe, Joanna C. Murrell, Siobhan Mullan
Summary: The awareness and use of canine quality of life (QOL) assessment tools in veterinary practice in the UK is low. Although most veterinary professionals are willing to use these tools, lack of time and potential resistance from owners are barriers to their use. This study suggests that QOL assessment tools are not well disseminated to veterinary professionals and that various barriers inhibit their use.
Article
Agriculture, Multidisciplinary
Andrew S. Cooke, Siobhan Mullan, Charlie Morten, Joanna Hockenhull, Phil Le-Grice, Kate Le Cocq, Michael R. F. Lee, Laura M. Cardenas, M. Jordana Rivero
Summary: Animal welfare encompasses all aspects of an animal's life and interactions. This study compared two beef cattle systems and their herds in terms of various indicators. The results showed that providing summer grazing to the cattle seemed to have welfare benefits, including more positive behavior and slightly better health indicators.
JOURNAL OF AGRICULTURAL SCIENCE
(2023)
Article
Veterinary Sciences
Rachel Annan, Leah E. Trigg, Jo Hockenhull, Kate Allen, Deborah Butler, Mathilde Valenchon, Siobhan Mullan
Summary: Racehorse welfare is a growing concern, but there is limited scientific evidence. This study aimed to assess racehorse welfare using objective methods. Thirteen training yards were visited and 353 horses were observed. The horses generally had good physical health, with 94% having an ideal body condition. The welfare assessment protocol used is suitable for collecting racehorse welfare data.
FRONTIERS IN VETERINARY SCIENCE
(2023)
Article
Veterinary Sciences
Anne Quain, Michael P. Ward, Siobhan Mullan
Summary: Veterinary team members encounter various ethical challenges in their work, which can negatively impact their well-being. A study analyzed published ethical vignettes from the veterinary literature and identified common types of ethical challenges, such as those involving dogs, livestock, and cattle. These findings contribute to a better understanding of the types of ethical challenges faced by veterinary team members and can inform training and preparation for navigating these challenges. Additionally, the study highlights factors contributing to these challenges and suggests potential solutions.
VETERINARY SCIENCES
(2022)