Filters








5,541 Hits in 8.5 sec

Evaluating bag-of-visual-words representations in scene classification

Jun Yang, Yu-Gang Jiang, Alexander G. Hauptmann, Chong-Wah Ngo
2007 Proceedings of the international workshop on Workshop on multimedia information retrieval - MIR '07  
Based on keypoints extracted as salient image patches, an image can be described as a "bag of visual words" and this representation has been used in scene classification.  ...  selection, to generate image representations that differ in the dimension, selection, and weighting of visual words.  ...  Zipf's law says that the frequency of any (visual) word is roughly inversely proportional to its rank in terms of frequency.  ... 
doi:10.1145/1290082.1290111 dblp:conf/mir/YangJHN07 fatcat:ov3j6mttdjcdhdeqltr3boepfa

Toward a higher-level visual representation for content-based image retrieval

Ismail El sayad, Jean Martinet, Thierry Urruty, Chabane Djeraba
2010 Multimedia tools and applications  
The aim of the approach is to build a meaningful mid-level representation of images to be used later on for matching between a query image and other images in the desired database.  ...  Thirdly, a new spatial weighting scheme is introduced that consists of weighting visual words according to the probability of each visual word to belong to each of the n Gaussian.  ...  Having these convolutions decreases the number of detected interest points and this contributes as a limitation for Fast Hessian in images with rare texture.  ... 
doi:10.1007/s11042-010-0596-x fatcat:lkbyc2z6frfu7alofppr3ydyqi

Toward a higher-level visual representation for content-based image retrieval

Ismail El sayad, Jean Martinet, Thierry Urruty, Samir Amir, Chabane Djeraba
2010 Proceedings of the 8th International Conference on Advances in Mobile Computing and Multimedia - MoMM '10  
The aim of the approach is to build a meaningful mid-level representation of images to be used later on for matching between a query image and other images in the desired database.  ...  Thirdly, a new spatial weighting scheme is introduced that consists of weighting visual words according to the probability of each visual word to belong to each of the n Gaussian.  ...  Having these convolutions decreases the number of detected interest points and this contributes as a limitation for Fast Hessian in images with rare texture.  ... 
doi:10.1145/1971519.1971556 dblp:conf/momm/ElsayadMUAD10 fatcat:pvudf2m2qbb75cjp2u2rxucxkq

Saliency moments for image categorization

Miriam Redi, Bernard Merialdo
2011 Proceedings of the 1st ACM International Conference on Multimedia Retrieval - ICMR '11  
Results show that this new type of image description outperforms the traditional global features on scene and object categorization, for a variety of challenging datasets.  ...  In this paper we present Saliency Moments, a new, holistic descriptor for image recognition inspired by two biological vision principles: the gist perception and the selective visual attention.  ...  ACKNOWLEDGMENTS The authors would like to thank Matteo Dell'Amico and Judith Redi for their precious comments and discussions. This Research was funded by Amadeus.  ... 
doi:10.1145/1991996.1992035 dblp:conf/mir/RediM11 fatcat:nzu5gloiqngirekypsyxauloua

From Text to Images: Weighting Schemes for Image Retrieval

Pierre Tirilly, Vincent Claveau, Patrick Gros
2015 Journal of Multimedia  
Bags of visual words are the most studied image description technique in the last years. This representation of images raised new possibilities as well as new research issues.  ...  It first shows that the optimal setting for distances and weighting schemes depends on the nature of the visual content of the images considered.  ...  frequency: It emphasizes the importance of frequent visual terms. 2) Global Weights: a) Inverse document frequency: Inverse document frequency (idf) is born from the observation that words occurring in  ... 
doi:10.4304/jmm.10.01.1-21 fatcat:vup6pwn32ffatjgrdhylghzzwq

Efficient Boosted Exemplar-Based Face Detection

Haoxiang Li, Zhe Lin, Jonathan Brandt, Xiaohui Shen, Gang Hua
2014 2014 IEEE Conference on Computer Vision and Pattern Recognition  
In contrast, exemplar-based face detection has been shown to be effective, even under these challenging conditions, primarily because a large exemplar database is leveraged to cover all possible visual  ...  However, relying heavily on a large exemplar database to deal with the face appearance variations makes the detector impractical due to the high space and time complexity.  ...  start-up funds from Stevens Institute of Technology, a Google Research Faculty Award, a gift grant from Microsoft Research, and a gift grant from NEC Labs America.  ... 
doi:10.1109/cvpr.2014.238 dblp:conf/cvpr/LiLBSH14 fatcat:h4zqizvqajaqhnh7xt2pefkbgq

Acoustic Inversion in Optoacoustic Tomography: A Review

Amir Rosenthal, Vasilis Ntziachristos, Daniel Razansky
2014 Current Medical Imaging Reviews  
From an imaging standpoint, the acoustic inverse problem relates to forming an image from the measured acoustic data, whereas the optical inverse problem relates to quantifying the formed image.  ...  The hybrid nature of optoacoustic tomography gives rise to two distinct inverse problems: The optical inverse problem, related to the propagation of the excitation light in tissue, and the acoustic inverse  ...  Since reconstructing a single slice of the optoacoustic source requires fewer projections than the entire 3D volume, cylindrically focused detectors enable fast data acquisition for selective-plane imaging  ... 
doi:10.2174/15734056113096660006 pmid:24772060 pmcid:PMC3996917 fatcat:qajc6lc3dfelxl3bzzfxbyfasm

Content-based visual search learned from social media

Xirong Li
2012 ACM SIGMultimedia Records  
Acknowledgements Being a PhD is a non-trivial trip. I would like to use this dedicated section to thank people who practically or mentally helped me accomplish the trip.  ...  It is my pleasure to work in the ISIS group: Michael, Ivo, and Jan as my ping pong mates for years, Victoria, Vladimir, Dung, and Stratios for sharing offices, Cor, Jan-Mark, and Theo for interesting discussions  ...  [125] which re-weights the frequency of a tag by its inverse document frequency (tf-idf).  ... 
doi:10.1145/2206765.2206774 fatcat:kxk6kciwhfe2hcqw546ez2coku

Visual Phrases for Exemplar Face Detection

Vijay Kumar, Anoop Namboodiri, C. V. Jawahar
2015 2015 IEEE International Conference on Computer Vision (ICCV)  
We discover visual phrases that contain semantically related visual words and exploit them for detection along with the visual words.  ...  Current schemes, however, make an assumption of independence between the visual words, ignoring their relations in the process. They also ignore the spatial consistency of the visual words.  ...  We achieve this using the concept of spatial consistency introduced earlier for visual words.  ... 
doi:10.1109/iccv.2015.231 dblp:conf/iccv/KumarNJ15 fatcat:g4mifm442veetaengcb2vndprm

Attribute-based People Search

Rogerio Feris, Russel Bobbitt, Lisa Brown, Sharath Pankanti
2014 Proceedings of International Conference on Multimedia Retrieval - ICMR '14  
Moreover, we show impressive results using images from the recent Boston marathon bombing event, where our system can rapidly retrieve the two suspects based on their attributes from a database containing  ...  The system we developed is capable of answering user queries such as "show me all people with a beard and sunglasses, wearing a white hat and a patterned blue shirt, from all metro cameras in the downtown  ...  The weighting is a product of two terms: the word frequency n id n d , and the inverse document frequency log N n i .  ... 
doi:10.1145/2578726.2578732 dblp:conf/mir/FerisBBP14 fatcat:uawezbik5nd3dmjtnac6am5vzq

A Review of Codebook Models in Patch-Based Visual Object Recognition

Amirthalingam Ramanan, Mahesan Niranjan
2011 Journal of Signal Processing Systems  
The discriminative power of such a visual codebook determines the quality of the codebook model, whereas the size of the codebook controls the complexity of the model.  ...  The key role of a visual codebook is to provide a way to map the low-level features into a fixed-length vector in histogram space to which standard classifiers can be directly applied.  ...  Acknowledgements The first author is supported in part by a grant from the University of Jaffna, Sri Lanka under the IRQUE Project funded by the World Bank and in part by a grant from the School of Electronics  ... 
doi:10.1007/s11265-011-0622-x fatcat:bowdhn7xrvbxfbt5k2ernaud3m

Detection and visualization of encoded local features as anatomical predictors in cross-sectional images of Lauraceae

Sung-Wook Hwang, Kayoko Kobayashi, Junji Sugiyama
2020 Journal of Wood Science  
In the bag-of-features framework with the SIFT features, each image is represented by a histogram of codewords.  ...  An analysis of the term frequency-inverse document frequency weights revealed that blob-based codewords are generally shared by all species, whereas corner-based codewords are more species specific.  ...  We thank Kimberly Moravec, PhD, from Edanz Group (http://www.edanz editi ng.com/ ac) for editing a draft of this manuscript. SH is a recipient of JSPS Postdoctoral Fellowship for Foreign Researchers.  ... 
doi:10.1186/s10086-020-01864-5 fatcat:mzxplur6uvhgdnb2vkco2fknam

Direct Intracranial, fMRI, and Lesion Evidence for the Causal Role of Left Inferotemporal Cortex in Reading

Raphaël Gaillard, Lionel Naccache, Philippe Pinel, Stéphane Clémenceau, Emmanuelle Volle, Dominique Hasboun, Sophie Dupont, Michel Baulac, Stanislas Dehaene, Claude Adam, Laurent Cohen
2006 Neuron  
Models of the "visual word form system" postulate that a left occipitotemporal region implements the automatic visual word recognition required for efficient reading.  ...  Prior to surgery, when reading was normal, fMRI revealed a normal mosaic of ventral visual selectivity for words, faces, houses, and tools.  ...  Supplemental Data The Supplemental Data for this article can be found online at http:// www.neuron.org/cgi/content/full/50/2/191/DC1/.  ... 
doi:10.1016/j.neuron.2006.03.031 pmid:16630832 fatcat:o5zogqwcobge7hzawtclbya73y

A Novel Language-Model-Based Approach for Image Object Mining and Re-ranking

Jen-Hao Hsiao, Chu-Song Chen, Ming-Syan Chen
2008 2008 Eighth IEEE International Conference on Data Mining  
One leading framework for image object mining is the bag-of-words (BOW) approach. The idea is to encode an image as a collection of visual words of the quantized local patches.  ...  Objects in the image can then be retrieved through inferring the semantic topics associated with the set of visual words.  ...  In text categorization, words with small term frequency are removed since rare words are relatively non-informative for category prediction.  ... 
doi:10.1109/icdm.2008.83 dblp:conf/icdm/HsiaoCC08 fatcat:zcrqmgx6pzfljmtf24x53vk3wm

Explaining Deep Convolutional Neural Networks on Music Classification [article]

Keunwoo Choi, George Fazekas, Mark Sandler
2016 arXiv   pre-print
We introduce auralisation of a CNN to understand its underlying mechanism, which is based on a deconvolution procedure introduced in [2].  ...  Auralisation of a CNN is converting the learned convolutional features that are obtained from deconvolution into audio signals.  ...  It is a similar result to the result that is often obtained in visual image recognition, where CNNs learn line detectors with various directions (also known as edge detectors), which are combined to create  ... 
arXiv:1607.02444v1 fatcat:dqmsi3pgezb6fnlc75s6tf77lq
« Previous Showing results 1 — 15 out of 5,541 results