Filters








20,585 Hits in 5.1 sec

Hierarchical Long-Term Learning for Automatic Image Annotation [chapter]

Donn Morrison, Stéphane Marchand-Maillet, Eric Bruno
Semantic Multimedia  
This paper introduces a hierarchical process for propagating image annotations throughout a partially labelled database.  ...  Long-term learning, where users' query and browsing patterns are retained over multiple sessions, is used to guide the propagation of keywords onto image regions based on low-level feature distances.  ...  Section 3 introduces our method for automatic annotation using regions of low-level features and long-term learning. Next, Section 4 details the image database we use and the experiments followed.  ... 
doi:10.1007/978-3-540-77051-0_3 dblp:conf/samt/MorrisonMB07 fatcat:k3plakxdarhtpc37tm22w2xpre

Hierarchical classification for automatic image annotation

Jianping Fan, Yuli Gao, Hangzai Luo
2007 Proceedings of the 30th annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '07  
In this paper, a hierarchical classification framework has been proposed for bridging the semantic gap effectively and achieving multi-level image annotation automatically.  ...  hierarchical image classifier training with automatic error recovery.  ...  CONCLUSIONS In this paper, we have proposed a novel algorithm for automatic multi-level image annotation via hierarchical classification.  ... 
doi:10.1145/1277741.1277763 dblp:conf/sigir/FanGL07 fatcat:glefia7ybzb7fh5ormjohd4zkq

Automatic Image Annotation with Relevance Feedback and Latent Semantic Analysis [chapter]

Donn Morrison, Stéphane Marchand-Maillet, Eric Bruno
2008 Lecture Notes in Computer Science  
We demonstrate how automatic annotation of images can be implemented on partially annotated databases by learning imageconcept relationships from positive examples via inter-query learning.  ...  The goal of this paper is to study the image-concept relationship as it pertains to image annotation.  ...  This method of exchanging RF instances and images for the documents and term vocabulary was also used in a later study where the authors use long-term learning in the PicSOM retrieval system [8] .  ... 
doi:10.1007/978-3-540-79860-6_6 fatcat:l5y7ly5jhrdxflr4r6v4dwhkqm

Measuring Fine-Grained Domain Relevance of Terms: A Hierarchical Core-Fringe Approach [article]

Jie Huang, Kevin Chen-Chuan Chang, Jinjun Xiong, Wen-mei Hwu
2021 arXiv   pre-print
To reduce expensive human efforts, we employ automatic annotation and hierarchical positive-unlabeled learning.  ...  To support a fine-grained domain without relying on a matching corpus for supervision, we develop hierarchical core-fringe learning, which learns core and fringe terms jointly in a semi-supervised manner  ...  Acknowledgments We thank the anonymous reviewers for their valuable comments and suggestions. This material is  ... 
arXiv:2105.13255v1 fatcat:mz5sxz7gtvhaljgjihwwit5ami

Photo annotation on a camera phone

Anita Wilhelm, Yuri Takhteyev, Risto Sarvas, Nancy Van House, Marc Davis
2004 Extended abstracts of the 2004 conference on Human factors and computing systems - CHI '04  
In this paper we present usability issues encountered in using a camera phone as an image annotation device immediately after image capture and users' responses to use of such a system.  ...  The system uses camera phones with a lightweight client application and a server to store the images and metadata and assists the user in annotation on the camera phone by providing guesses about the content  ...  As image quality improves, we expect that users will add to these ad hoc uses more traditional (long-term) imaging behavior with more need for metadata.  ... 
doi:10.1145/985921.986075 dblp:conf/chi/WilhelmTSHD04 fatcat:as52ic233bbotifncf5xhnogvm

Attentive Hierarchical Label Sharing for Enhanced Garment and Attribute Classification of Fashion Imagery

Stefanos-Iordanis Papadopoulos, Christos Koutlis, Manjunath Sudheer, Martina Pugliese, Delphine Rabiller, Symeon Papadopoulos, Ioannis Kompatsiaris
2021 Zenodo  
Our approach enables progressively focusing on appropriately detailed features for automatically learning the hierarchical relations of fashion and enabling predictions on images with complete outfits.  ...  For the category and attribute-level classification stages we examine a hierarchical label sharing (HLS) technique in two settings: (1) single-task learning (STL w/ HLS) and (2) multi-task learning with  ...  We would also like to thank Jamie Sutherland from Mallzee for his thoughts and input in this work.  ... 
doi:10.5281/zenodo.5566645 fatcat:2d7splyzivfinpgfhw4tzs46sy

Multimodal Recurrent Model with Attention for Automated Radiology Report Generation [chapter]

Yuan Xue, Tao Xu, L. Rodney Long, Zhiyun Xue, Sameer Antani, George R. Thoma, Xiaolei Huang
2018 Lecture Notes in Computer Science  
The proposed model incorporates the Convolutional Neural Networks (CNNs) with the Long Short-Term Memory (LSTM) in a recurrent way.  ...  Furthermore, generating detailed paragraph descriptions for medical images remains a challenging problem.  ...  [18] proposed a deep learning framework to automatically annotate chest x-rays with Medical Subject Headings (MeSH) annotations for the first time.  ... 
doi:10.1007/978-3-030-00928-1_52 fatcat:hxsh7dto6zfpfcipjwxa4bvip4

Fusing Body Posture with Facial Expressions for Joint Recognition of Affect in Child-Robot Interaction [article]

Panagiotis P. Filntisis, Niki Efthymiou, Petros Koutras, Gerasimos Potamianos, Petros Maragos
2019 arXiv   pre-print
Towards this goal we propose a method for automatic recognition of affect that leverages body expressions alongside facial expressions, as opposed to traditional methods that usually focus only on the  ...  both the individual modalities, as well as for the whole body emotion.  ...  [43] used a hierarchical combination of Bidirectional long short-term memory (LSTM) and convolutional layers with body-face fusion using support vector machines. Bänziger et al.  ... 
arXiv:1901.01805v2 fatcat:4ylt2boqibeunkzqexx75kkoyi

Human Pose Co-Estimation and Applications

M. Eichner, V. Ferrari
2012 IEEE Transactions on Pattern Analysis and Machine Intelligence  
We show that PCE leads to better pose estimation in such images, and it learns meaningful prototypes which can be used as priors for pose estimation in novel images.  ...  The second application is learning prototype poses characterizing a pose class directly from an image search engine queried by the class name (e.g. 'lotus pose').  ...  We have demonstrated its benefits for estimating poses in images of synchronized activities and for learning prototypes of pose classes fully automatically, directly by querying Google Images with the  ... 
doi:10.1109/tpami.2012.85 pmid:22487983 fatcat:l256cfttizc3ta6mj5tjic4gia

AutoMPR: Automatic detection of standard planes in 3D echocardiography

Xiaoguang Lu, Bogdan Georgescu, Yefeng Zheng, Joanne Otsuki, Dorin Comaniciu
2008 2008 5th IEEE International Symposium on Biomedical Imaging: From Nano to Macro  
It provides for evaluation a more complete heart representation in comparison to conventional 2D echocardiography.  ...  3D echocardiography is one of the emerging real-time imaging modalities that is increasingly used in clinical practice to assess cardiac function.  ...  LV long axis and the MPR (for SAXB and SAXA).  ... 
doi:10.1109/isbi.2008.4541237 dblp:conf/isbi/LuGZOC08 fatcat:kvsqtowkuvfqnouqqaqpdkbn7q

Image Retrieval: Modelling Keywords via Low-level Features

Zenonas Theodosiou
2015 ELCVIA Electronic Letters on Computer Vision and Image Analysis  
It discusses and presents several studies referring to: (a) low-level feature extraction and selection for the task of automatic annotation of images, (b) training algorithms that can be utilized for keyword  ...  In recent years, much effort has been invested on automatic image annotation methods, since the manual assignment of keywords (which is necessary for text-based image retrieval) is a time consuming and  ...  Although idea of automatic image annotation by keywords modelling is very promising, there is a long way to go before this method can be utilized to solve the problem of automatic image annotation.  ... 
doi:10.5565/rev/elcvia.725 fatcat:zbpgc3jstnd5nf5ltcly3rdbhi

Hierarchical Memory Learning for Fine-Grained Scene Graph Generation [article]

Youming Deng, Yansheng Li, Yongjun Zhang, Xiang Xiang, Jian Wang, Jingdong Chen, Jiayi Ma
2022 arXiv   pre-print
In order to alleviate the negative impact of the suboptimum mixed-granularity annotation and long-tail effect problems, this paper proposes a novel Hierarchical Memory Learning (HML) framework to learn  ...  In order to realize this hierarchical learning pattern, this paper, for the first time, formulates the HML framework using the new Concept Reconstruction (CR) and Model Reconstruction (MR) constraints.  ...  Long-Tail Learning Only a few works like [7, 10, 17, 29] cast importance on the long-tail effect in VG. In fact, many long-tail learning strategies can be used in SGG.  ... 
arXiv:2203.06907v3 fatcat:4xxgw4zonjhypfg3vvdecbevvy

Toward Interactive Self-Annotation For Video Object Bounding Box: Recurrent Self-Learning And Hierarchical Annotation Based Framework

Trung-Nghia Le, Sugimoto Akihiro, Shintaro Ono, Hiroshi Kawasaki
2020 2020 IEEE Winter Conference on Applications of Computer Vision (WACV)  
To this end, we propose a novel Hierarchical Correction module, where the annotated frame-distance binarizedly decreases at each time step, to utilize the strength of CNN for neighbor frames.  ...  Our method is based on recurrent self-supervised learning and consists of two processes: automatic process and interactive process, where the automatic process aims to build a supported detector to speed  ...  LableME [28] is a popular web-based tool for annotating arbitrary shapes of an object, which has two versions: LableME-Image (2008) [28] for only image annotation, and LableME-Video (2009) [43] for  ... 
doi:10.1109/wacv45572.2020.9093398 dblp:conf/wacv/LeSOK20 fatcat:gihnxvt4znhv3f23h5bj2hrnii

PadChest: A large chest x-ray image dataset with multi-label annotated reports [article]

Aurelia Bustos, Antonio Pertusa, Jose-Maria Salinas, Maria de la Iglesia-Vayá
2019 arXiv   pre-print
We present a labeled large-scale, high resolution chest x-ray dataset for the automated exploration of medical images along with their associated reports.  ...  position views and additional information on image acquisition and patient demography.  ...  The Medical Image Bank of the Valencian Community as well as de-identification and anonymization services, were partially funded by the Regional Ministry of Health (FEDER program) and the Horizon 2020  ... 
arXiv:1901.07441v2 fatcat:uuhka6akyrhr7orlppbgymxjsy

Fast Object Class Labelling via Speech [article]

Michael Gygli, Vittorio Ferrari
2019 arXiv   pre-print
Modern approaches rely on a hierarchical organization of the vocabulary to reduce annotation time, but remain expensive (several minutes per image for the 200 classes in ILSVRC).  ...  As additional advantages, annotators can simultaneously speak and scan the image for objects, the interface can be kept extremely simple, and using it requires less mouse movement.  ...  learn to use the provided vocabulary for naming objects with high fidelity. • Analyse the accuracy of models for automatic speech recognition (ASR) and show that it supports deriving high-quality annotations  ... 
arXiv:1811.09461v2 fatcat:o7aj7mkftnfoxepyhnw55moaky
« Previous Showing results 1 — 15 out of 20,585 results