Article
Agriculture, Multidisciplinary
Xing Wang, Hanwen Kang, Hongyu Zhou, Wesley Au, Chao Chen
Summary: Field robotic harvesting is a promising technique in agricultural industry. This study proposes a geometry-aware network, A3N, and a global-to-local scanning strategy to enable robots to accurately recognize and retrieve fruits in complex field environments.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2022)
Article
Computer Science, Information Systems
Eleni Vrochidou, George K. Sidiropoulos, Athanasios G. Ouzounis, Anastasia Lampoglou, Ioannis Tsimperidis, George A. Papakostas, Ilias T. Sarafis, Vassilis Kalpakis, Andreas Stamkos
Summary: This work provides a comprehensive study on marble crack segmentation using deep learning techniques. The authors propose efficient network architectures and feature extraction methods, making an important contribution to addressing the problem of marble crack segmentation.
Article
Food Science & Technology
L. G. Divyanth, Peeyush Soni, Chaitanya Madhaw Pareek, Rajendra Machavaram, Mohammad Nadimi, Jitendra Paliwal
Summary: This study proposes a deep learning-based Faster R-CNN model with an attention mechanism to detect coconut clusters. The model achieved high accuracy and robustness in the tests, providing a foundation for developing a complete vision system for coconut harvesting.
Article
Environmental Sciences
Tao Li, Qingchun Feng, Quan Qiu, Feng Xie, Chunjiang Zhao
Summary: A novel 3D fruit localization method based on deep learning and a new frustum-based point-cloud-processing method is proposed in this paper, and experiments demonstrate its effectiveness, with significantly improved accuracy and performance compared to conventional methods.
Article
Plant Sciences
Xinzhao Zhou, Xiangjun Zou, Wei Tang, Zhiwei Yan, Hewei Meng, Xiwen Luo
Summary: In this study, a novel algorithm was proposed for unstructured road extraction and roadside fruit synchronous recognition in complex orchard environments. A preprocessing method tailored to field orchards was proposed to reduce the interference of adverse factors, and a road region extraction method based on dual-space fusion was developed. A fusion recognition framework was established, employing the optimized YOLOv7 model for roadside fruit detection. The proposed method demonstrated improved road extraction and increased fruit identification accuracy and speed.
FRONTIERS IN PLANT SCIENCE
(2023)
Article
Computer Science, Artificial Intelligence
Shuai Luo, Yujie Li, Pengxiang Gao, Yichuan Wang, Seiichi Serikawa
Summary: This paper reviews state-of-the-art image segmentation methods based on meta-learning, introducing the background and differences with other similar methods, discussing various types of meta-learning methods and their applications in image segmentation, conducting experimental comparisons, and highlighting future trends of meta-learning in image segmentation.
PATTERN RECOGNITION
(2022)
Article
Engineering, Electrical & Electronic
Lemiao Yang, Fuqiang Zhou, Lin Wang
Summary: This article proposes a scratch detection method combining deep learning and image segmentation algorithm, which has advantages in both accuracy and detection speed of scratch recognition, and can effectively segment scratch pixels.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2022)
Article
Computer Science, Artificial Intelligence
Yunchao Tang, Hao Zhou, Hongjun Wang, Yunqi Zhang
Summary: In this study, a fruit detection model method called YOLO-Oleifera was developed based on the YOLOv4-tiny model. The YOLO-Oleifera model improved upon the YOLOv4-tiny model by adding convolutional kernels and using the k-means++ clustering algorithm. It innovatively used bounding boxes for stereo matching to determine the picking position of the fruit. Experimental results showed that the model achieved stable detection performance and low computational complexity under different illumination conditions.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Environmental Sciences
Amritha Haridasan, Jeena Thomas, Ebin Deni Raj
Summary: Automatic detection and analysis of rice crop diseases is essential in the farming industry to prevent resource wastage, reduce yield losses, and improve treatment efficiency. A computer vision-based approach using image processing, machine learning, and deep learning techniques is proposed to accurately detect and classify rice plant diseases. Five primary diseases that frequently affect Indian rice fields can be identified through image segmentation and visual content recognition. The suggested deep learning-based strategy achieved a high validation accuracy of 0.9145 using a support vector machine classifier and convolutional neural networks. After recognition, a predictive remedy is recommended to assist agriculture-related individuals and organizations in combating these diseases.
ENVIRONMENTAL MONITORING AND ASSESSMENT
(2023)
Article
Agricultural Engineering
Omeed Mirbod, Daeun Choi, Paul H. Heinemann, Richard P. Marini, Long He
Summary: A machine vision system combined with deep neural network models is developed to accurately measure the size of apple fruit, which can be used for orchard management decision-making, fruit quality assessment, and yield prediction. This technology can handle occluded fruit and improve the accuracy of fruit size measurement.
BIOSYSTEMS ENGINEERING
(2023)
Article
Computer Science, Information Systems
Mazen Mushabab Alqahtani, Ashit Kumar Dutta, Sultan Almotairi, M. Ilayaraja, Amani Abdulrahman Albraikan, Fahd N. Al-Wesabi, Mesfer Al Duhayyim
Summary: Recent advancements in digital cameras and electronic gadgets have led to the development of automated apple leaf disease detection models based on Machine Learning and Deep Learning techniques, which are viable alternatives to traditional visual inspection models. This paper proposes an ESFO-EALD model, utilizing Effective Sailfish Optimizer and EfficientNet, for automatic detection of plant leaf diseases. Median Filtering is employed to enhance the quality of apple plant leaf images, while SFO with Kapur's entropy-based segmentation technique is used to identify the affected plant region. Adam optimizer, EfficientNet-based feature extraction, and Spiking Neural Network classification are employed for detecting and classifying apple plant leaf images. Extensive simulations validate the effectiveness of the ESFO-EALD technique on benchmark datasets, showing its superiority over existing approaches.
CMC-COMPUTERS MATERIALS & CONTINUA
(2023)
Article
Computer Science, Information Systems
Francisco E. Fernandes jr, Luis Gustavo Nonato, Jo Ueyama
Summary: Floods cause significant economic and social losses, but people in developing countries often lack access to expensive flooding alert systems. To address this, the authors propose a cheap and robust River Flooding Detection System that utilizes raw images from off-the-shelf cameras without preprocessing. The system accurately measures river levels using semantic segmentation and computer vision techniques, and automatically sends alerts when water levels reach or exceed dangerous thresholds.
MULTIMEDIA TOOLS AND APPLICATIONS
(2022)
Article
Plant Sciences
Jianqiang Lu, Ruifan Yang, Chaoran Yu, Jiahan Lin, Wadi Chen, Haiwei Wu, Xin Chen, Yubin Lan, Weixing Wang
Summary: This study proposes a citrus green fruit detection method based on improved Mask-RCNN, which can effectively improve the detection accuracy through deep learning technology and is of great significance for the intelligent production of citrus.
FRONTIERS IN PLANT SCIENCE
(2022)
Review
Engineering, Mechanical
Meenakshi Suresh Kumar, Santhakumar Mohan
Summary: The progressive application of multidisciplinary research and development in agriculture is driving the automation evolution in various subsectors. This paper provides a brief analysis of selective fruit harvesting techniques from 2017 to 2022, highlighting the importance of accurate information capture, challenges in fruit detection, and the need for improved fruit grasping and detachment for quality preservation and increased production rate.
PROCEEDINGS OF THE INSTITUTION OF MECHANICAL ENGINEERS PART C-JOURNAL OF MECHANICAL ENGINEERING SCIENCE
(2023)
Review
Agronomy
Feng Xiao, Haibin Wang, Yueqin Xu, Ruiqing Zhang
Summary: Continuing progress in machine learning has led to significant advancements in fruit detection and automatic harvesting using deep learning. This paper provides a comprehensive overview and review of fruit detection and recognition based on deep learning from 2018 up to now, focusing on current challenges and proposing feasible solutions and future development trends. It aims to serve as a reference for follow-up research in this field.
Article
Computer Science, Artificial Intelligence
Luan Tran, Xi Yin, Xiaoming Liu
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2019)
Article
Computer Science, Artificial Intelligence
Mehdi Bahri, Eimear O' Sullivan, Shunwang Gong, Feng Liu, Xiaoming Liu, Michael M. Bronstein, Stefanos Zafeiriou
Summary: This paper introduces a new learning-based approach for non-rigid registration of face scans, which is faster, more robust, has fewer parameters, and can generalize to previously unseen datasets compared to standard registration algorithms. The model's registration quality is extensively evaluated on diverse data, demonstrating robustness and generalizability across different facial scans.
INTERNATIONAL JOURNAL OF COMPUTER VISION
(2021)
Article
Automation & Control Systems
Kaixiang Zhang, Kyle Lammers, Pengyu Chu, Zhaojian Li, Renfu Lu
Summary: This study presents a robotic apple harvesting prototype with mechatronic design and motion control. The prototype utilizes deep learning for fruit detection and localization, incorporates a pneumatic/motor actuation mechanism for dexterous movements, and features a vacuum-based end-effector for apple detachment. Additionally, a nonlinear control scheme is developed for accurate and agile motion control, demonstrated through field experiments to showcase the robot's performance in apple harvesting.
Proceedings Paper
Computer Science, Artificial Intelligence
Armand Comas, Tim K. Marks, Hassan Mansour, Suhas Lohit, Yechi Ma, Xiaoming Liu
Summary: Imaging photoplethysmography (iPPG) is a method used to estimate a person's pulse waveform by processing a video of their face, and in situations with insufficient visible spectrum illumination, a modular framework with a novel time-series U-net architecture can be used for heartbeat signal estimation. The proposed method outperforms existing models on challenging datasets containing monochromatic NIR videos taken in different conditions.
2021 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING (ICIP)
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Yunfei Long, Daniel Morris, Xiaoming Liu, Marcos Castro, Punarjay Chakravarty, Praveen Narayanan
Summary: This paper presents a closed-form solution for the full-velocity estimate of Doppler returns using optical flow from camera images, and addresses the association problem between radar returns and camera images with a trained neural network. Experimental results on the nuScenes dataset validate the effectiveness of the method in velocity estimation and accumulation of radar points, showing significant improvements over the state-of-the-art.
2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2021)
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Feng Liu, Luan Tran, Xiaoming Liu
Summary: The study focuses on inferring the 3D structure of a generic object from a 2D image. By utilizing semi-supervised learning and decomposing the image into latent representations, the approach enables modeling and model fitting using real 2D images, resulting in superior 3D reconstruction.
2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Ziqian Bai, Zhaopeng Cui, Xiaoming Liu, Ping Tan
Summary: This paper introduces a method for riggable 3D face reconstruction from monocular images, utilizing a trainable network to estimate personalized face rig and per-image parameters, achieving beyond static reconstructions and supporting downstream applications such as video retargeting. The network utilizes in-network optimization to enforce constraints and data-driven priors to constrain the ill-posed monocular setting, leading to state-of-the-art reconstruction accuracy, robustness, and generalization ability.
2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Saif Imran, Xiaoming Liu, Daniel Morris
Summary: The method proposed in the paper models both foreground and background depths in difficult occlusion-boundary regions by using a multi-hypothesis depth representation. It performs twin-surface extrapolation instead of interpolation in these regions. The approach trains a network to simultaneously do surface extrapolation and surface fusion using an asymmetric loss function on a novel twin-surface representation.
2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Andrew Hou, Ze Zhang, Michel Sarkis, Ning Bi, Yiying Tong, Xiaoming Liu
Summary: This study introduces a novel deep face relighting method that can accurately handle shadows while maintaining local facial details. By predicting the ratio image between source and target images and modifying shadows using shadow masks, the method demonstrates state-of-the-art face relighting performance.
2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Yunfei Long, Daniel Morris, Xiaoming Liu, Marcos Castro, Punarjay Chakravarty, Praveen Narayanan
Summary: This study proposes a mapping method from radar returns to pixels to achieve image-guided radar and video depth completion. By integrating radar and video data at the pixel level, superior performance to using camera and radar alone is demonstrated.
2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021
(2021)
Article
Computer Science, Artificial Intelligence
Luan Tran, Xiaoming Liu
Summary: This paper proposes an innovative framework to learn a nonlinear 3DMM model from a large set of in-the-wild face images, significantly enhancing its representation power and making significant contributions to face alignment, 3D reconstruction, and face editing.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Ziqian Bai, Zhaopeng Cui, Jamal Ahmed Rahim, Xiaoming Liu, Ping Tan
2020 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR)
(2020)
Proceedings Paper
Computer Science, Artificial Intelligence
Luan Tran, Kihyuk Sohn, Xiang Yu, Xiaoming Liu, Manmohan Chandraker
2019 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2019)
(2019)