Article
Computer Science, Information Systems
Zhengrui Peng, Xinyi Gong, Bengang Wei, Xiangyi Xu, Shixiong Meng
Summary: This paper introduces an unsupervised learning method based on self-feature comparison for accurately locating and segmenting anomalies in fabric texture images. Compared to traditional methods, this approach performs better in locating anomalies on fiber texture surfaces.
Article
Engineering, Civil
Yubin Guo, Xinlei Qi, Jin Xie, Cheng-Zhong Xu, Hui Kong
Summary: In this paper, an unsupervised visible light-guided cross-spectrum depth-estimation framework is proposed. It achieves reliable depth maps under variant-illumination conditions with a pair of dual-spectrum images. Through training a depth-estimation base network, transferring features from the TIR domain to the VIS domain, and introducing a mechanism of cross-spectrum depth cycle-consistency, our method outperforms existing methods in depth estimation.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Xin Zhang, Ziliang Feng
Summary: This paper studies an unsupervised approach to person reidentification, using a clustering algorithm and contrastive learning to generate pseudolabels, and proposes a quantitative random selection strategy for cluster feature representation. Extensive experiments show that this method achieves state-of-the-art performance in unsupervised person re-ID tasks.
APPLIED INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Tao Wu, Wenzhuo Fan, Shuxian Li, Qingqing Li, Jianlin Zhang, Meihui Li
Summary: In this article, an unsupervised model is proposed to extract landmarks of objects in images. The authors combine Transformer structure with convolutional neural network structure to represent and encode the landmarks. Positive and negative sample pairs between landmarks are constructed to pull semantically consistent landmarks closer and push semantically inconsistent landmarks farther in the feature space. The proposed model achieves better performance than other unsupervised methods on various datasets.
IET COMPUTER VISION
(2023)
Article
Engineering, Environmental
Chuangchuang Zhou, Wouter Sterkens, Dillam Jossue Diaz-Romero, Isiah Zaplana, Jef Peeters
Summary: Recent developments in robotic demanufacturing have the potential to enhance the efficiency of recycling and resource recovery from WEEE. To achieve industrial adoption, a generic retrieval system called YODO was developed, using content-based image retrieval (CBIR) to identify product models and retrieve model-specific demanufacturing instructions. The system compares visual features of WEEE images with a database to find matches and demonstrates high performance in a laptop model identification case study. YODO showed a top-1 retrieval mean average precision (mAP) of 93.75%, learned 1079 unique product models, and achieved an 85% chance of the next laptop being registered in the database.
RESOURCES CONSERVATION AND RECYCLING
(2023)
Article
Agriculture, Multidisciplinary
Jehan-Antoine Vayssade, Xavier Godard, Mathieu Bonneau
Summary: Computer vision is an interesting tool for animal behavior monitoring because it limits animal handling and can be used to record various traits using only one sensor. However, it remains challenging to collect individual information, not only detecting animals and behavior, but also identifying them.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2023)
Article
Chemistry, Analytical
Gemma Canet Tarres, Montse Pardas
Summary: Foreground object segmentation is a crucial step for surveillance systems based on video sensor networks. Existing methods either use statistical background modeling or convolutional neural networks (CNNs), but the latter usually requires specific training for each scene. This study proposes a method that does not require scene-specific training by using statistical techniques to generate a rough mask and refining it using a network. The results obtained demonstrate improved performance compared to non-CNN methods and are among the best for context-unsupervised CNN systems.
Article
Environmental Sciences
Xupei Zhang, Hanlin Qin, Yue Yu, Xiang Yan, Shanglin Yang, Guanghao Wang
Summary: This paper presents a novel unsupervised low-light image enhancement method that utilizes frequency-domain features to achieve dynamic range adjustment and enhancement of low-light images. Experimental results demonstrate that the method performs well among unsupervised methods and approaches the performance level of supervised methods.
Article
Computer Science, Artificial Intelligence
Adrian Rosello, Jose J. Valero-Mas, Antonio Javier Gallego, Javier Saez-Perez, Jorge Calvo-Zaragoza
Summary: The use of deep learning in computer vision tasks can achieve remarkable results, but it depends on the availability of training data and its relationship with the application scenario. Domain adaptation techniques are crucial in robotics, where there is limited access to targeted environment data. To facilitate research in this area, Kurcuma provides a collection of datasets for kitchen utensil recognition, along with a baseline using domain-adversarial training.
PATTERN ANALYSIS AND APPLICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Xin Zhang, Keren Fu, Yanci Zhang
Summary: This paper focuses on unsupervised person re-identification and proposes a graph correlation module for improving centroid quality and updates centroids using original features, achieving superior results compared to other methods.
SIGNAL IMAGE AND VIDEO PROCESSING
(2023)
Article
Computer Science, Artificial Intelligence
S. Sridhar Raj, Munaga V. N. K. Prasad, Ramadoss Balakrishnan
Summary: An unsupervised spatial segmented clustering model (SSC-DRTD) is proposed to handle occluded person re-identification images, showing improved performance over state-of-the-art methods on benchmark datasets.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Automation & Control Systems
Hohyun Sim, Hyeonjoong Cho, Hankyu Lee
Summary: This study proposes the first nonparametric clustering algorithm for unsupervised temporal sign segmentation and identifies that traditional metrics do not sufficiently address over-segmentation.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2024)
Article
Computer Science, Artificial Intelligence
Giang Truong, Huu Le, Erchuan Zhang, David Suter, Syed Zulqarnain Gilani
Summary: This paper introduces a novel unsupervised learning framework that can solve robust model fitting problems directly without labeled data. The method is agnostic to input features and can be applied to various LP-type problems. Empirical results show that it outperforms existing supervised and unsupervised learning approaches and achieves competitive results compared to traditional methods.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2023)
Article
Engineering, Biomedical
Mingue Song, Yanggon Kim
Summary: This study proposes an unsupervised learning method based on an encoder-decoder scheme for automatic classification of ultrasound breast tumors, demonstrating its potential in enhancing discriminant capability. By introducing the sequential form of autoencoder and changing the target mapping of the object, the study successfully achieved effective classification without the need for annotated information.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2022)
Review
Chemistry, Analytical
Henri Hoyez, Cedric Schockaert, Jason Rambach, Bruno Mirbach, Didier Stricker
Summary: This article introduces the supervised and unsupervised methods of image-to-image translation and their advantages and disadvantages. It also classifies and revises the current state-of-the-art methods, and conducts a quantitative evaluation of these methods.
Article
Mathematics, Applied
Andrea Caponnetto, Yuan Yao
ANALYSIS AND APPLICATIONS
(2010)
Article
Mathematics, Applied
Ming Li, Andrea Caponnetto
ANALYSIS AND APPLICATIONS
(2011)
Article
Mathematics, Applied
Ernesto De Vito, Lorenzo Rosasco, Alessandro Toigo
APPLIED AND COMPUTATIONAL HARMONIC ANALYSIS
(2014)
Article
Operations Research & Management Science
Silvia Villa, Lorenzo Rosasco, Sofia Mosci, Alessandro Verri
COMPUTATIONAL OPTIMIZATION AND APPLICATIONS
(2014)
Article
Biotechnology & Applied Microbiology
Paolo Fardin, Andrea Cornero, Annalisa Barla, Sofia Mosci, Massimo Acquaviva, Lorenzo Rosasco, Claudio Gambini, Alessandro Verri, Luigi Varesio
JOURNAL OF BIOMEDICINE AND BIOTECHNOLOGY
(2010)
Article
Computer Science, Artificial Intelligence
Luca Baldassarre, Lorenzo Rosasco, Annalisa Barla, Alessandro Verri
Article
Biochemistry & Molecular Biology
Paolo Fardin, Annalisa Barla, Sofia Mosci, Lorenzo Rosasco, Alessandro Verri, Rogier Versteeg, Huib N. Caron, Jan J. Molenaar, Ingrid Ora, Alessandra Eva, Maura Puppo, Luigi Varesio
Article
Operations Research & Management Science
S. Villa, L. Rosasco, S. Mosci, A. Verri
Article
Robotics
Federico Ceola, Elisa Maiettini, Giulia Pasquale, Giacomo Meanti, Lorenzo Rosasco, Lorenzo Natale
Summary: In this study, we focus on designing a fast instance segmentation learning pipeline for robotic applications. The pipeline utilizes a pre-trained CNN for feature extraction and fast-to-train Kernel-based classifiers to adapt to the presence of novel objects or different domains. Additionally, a training protocol is proposed to shorten the training time.
IEEE TRANSACTIONS ON ROBOTICS
(2022)
Article
Mathematics, Applied
Francesca Bartolucci, Ernesto De Vito, Lorenzo Rosasco, Stefano Vigogna
Summary: This paper discusses the characterization of function spaces corresponding to neural networks using the theory of reproducing kernel Banach spaces and proves a representation theorem for a wide class of such spaces. Additionally, it characterizes the norm of a certain class of ReLU activation functions in terms of the inverse Radon transform of a bounded real measure.
APPLIED AND COMPUTATIONAL HARMONIC ANALYSIS
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Sean Ryan Fanello, Carlo Ciliberto, Matteo Santoro, Lorenzo Natale, Giorgio Metta, Lorenzo Rosasco, Francesca Odone
2013 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION WORKSHOPS (CVPRW)
(2013)
Article
Automation & Control Systems
Andrea Tacchetti, Pavan K. Mallapragada, Matteo Santoro, Lorenzo Rosasco
JOURNAL OF MACHINE LEARNING RESEARCH
(2013)
Article
Automation & Control Systems
Lorenzo Rosasco, Silvia Villa, Sofia Mosci, Matteo Santoro, Alessandro Verri
JOURNAL OF MACHINE LEARNING RESEARCH
(2013)
Article
Automation & Control Systems
Lorenzo Rosasco, Mikhail Belkin, Ernesto De Vito
JOURNAL OF MACHINE LEARNING RESEARCH
(2010)
