10 Hits in 3.9 sec

Comparative Evaluation of Hand-Crafted Image Descriptors vs. Off-the-Shelf CNN-Based Features for Colour Texture Classification under Ideal and Realistic Conditions

Raquel Bello-Cerezo, Francesco Bianconi, Francesco Di Maria, Paolo Napoletano, Fabrizio Smeraldi
2019 Applied Sciences  
Nonetheless, the impact of this new paradigm on the classification of fine-grained images—such as colour textures—is still controversial.  ...  On average, the results indicate a marked superiority of deep networks, particularly with non-stationary textures and in the presence of multiple changes in the acquisition conditions.  ...  LCC is the probability distribution (histogram) of the angle between the colour vector of the central pixel in the neighbourhood and the average colour vector of the peripheral pixels.  ... 
doi:10.3390/app9040738 fatcat:w3dzz267ljhgrl3tjrya7etzyq

Representation Learning for Fine-Grained Change Detection

Niall O'Mahony, Sean Campbell, Lenka Krpalkova, Anderson Carvalho, Joseph Walsh, Daniel Riordan
2021 Sensors  
This review focuses on the state-of-the-art methods, applications, and challenges of representation learning for fine-grained change detection.  ...  Our research focuses on methods of harnessing the latent metric space of representation learning techniques as an interim output for hybrid human-machine intelligence.  ...  In ordinal classification/ranking, the degree of similarity can be considered based on the ordering of the classes in R.  ... 
doi:10.3390/s21134486 pmid:34209075 fatcat:2shbcyvutvfbhhqrsmjipjdrzi

Self-Supervised Representation Learning: Introduction, Advances and Challenges [article]

Linus Ericsson, Henry Gouk, Chen Change Loy, Timothy M. Hospedales
2021 arXiv   pre-print
Self-supervised representation learning methods aim to provide powerful deep feature learning without the requirement of large annotated datasets, thus alleviating the annotation bottleneck that is one  ...  These methods have advanced rapidly in recent years, with their efficacy approaching and sometimes surpassing fully supervised pre-training alternatives across a variety of data modalities including image  ...  Clustering This family of methods focuses on dividing the training data into a number of groups with high intra-group similarity and low inter-group similarity.  ... 
arXiv:2110.09327v1 fatcat:qoprtdh4rzg6lcylgn5rafubpe

Understanding and advancing PDE-based image compression [article]

Pascal Tobias Peter, Universität Des Saarlandes, Universität Des Saarlandes
Finally, we propose a new framework for diffusion-based image colourisation that we use to build an efficient codec for colour images.  ...  These methods rely on sparse representations by means of a discrete cosine or wavelet transform.  ...  [176] , we classify colourisation methods as strokeor example-based.  ... 
doi:10.22028/d291-26652 fatcat:2d4waqj42nezzdrwhdfl2jdraq

Efficient image duplicate detection based on image analysis

Yannick Maret
Indeed, each original image known to the system is associated with an adapted binary detector, based on a support vector classifier, that estimates the probability that a test image is one of its duplicate  ...  The proposed content-based duplicate detection system classifies a test image by associating it with a label that corresponds to one of the original known images.  ...  The equalised intensity is used to classify a pixel into one of the three achromatic classes. The logarithmic Hue is used to classify a pixel in one of the seven chromatic classes.  ... 
doi:10.5075/epfl-thesis-3797 fatcat:3vfqierq3vbhbegeylpek2czeq

An accurate 3D human face model reconstruction scheme

Wenlong Wang, Duncan Gillies
This method allowed both the global trends and the local subtle features to be preserved and well represented.  ...  With current 3D shape model reconstruction using statistical modelling, the accuracy of the reconstruction is highly dependent on the nature of the source data and the behaviour of the feature extraction  ...  of neighbour pixels in different neighbourhoods.  ... 
doi:10.25560/32518 fatcat:oshnok3tbbgbdbsjnblteacpqe

Visual articulated tracking in cluttered environments [article]

Christian Rauch, University Of Edinburgh, University Of Edinburgh, Maurice Fallon, Timothy Hospedales
These contributions are evaluated on two robotic platforms in dif [...]  ...  Traditionally, robot state estimation has relied on proprioceptive sensors as the single source of information about the internal state.  ...  Conceptually, a feature vector only relates to visual properties of a pixel and its neighbourhood and similar properties result in very similar feature vectors.  ... 
doi:10.7488/era/458 fatcat:2ti7r3ou7vgfvonmdvx4mns3rm

Crowd-sourced reconstruction of building interiors [article]

Michael Peter, Universität Stuttgart, Universität Stuttgart
are based either on fingerprint maps of electro-magnetic signals (e.g.  ...  To overcome the flaws of these methods, they are often supported by models for the human movement which in turn rely on indoor maps.  ...  Acknowledgements Herewith, I would like to thank everybody who had an influence on the formation of this thesis.  ... 
doi:10.18419/opus-8729 fatcat:22vvy6be4fda5cyjr4jxxdtmf4

Medical image synthesis in the diagnosis and study of neurodegenerative diseases

Christopher Bowles, Daniel Reuckert, Engineering And Physical Sciences Research Council
One such family of techniques involves generating synthetic medical images.  ...  This thesis examines the use of medical image synthesis with a particular focus on applications in neurodegenerative diseases.  ...  Choice of similarity metric for patch-based methods: The majority of the patch based approaches to synthesis require the choice of some sort of similarity metric.  ... 
doi:10.25560/74457 fatcat:qc25mwojobb37pfijvwbtyyr4e

From differential equations to differential geometry : aspects of regularisation in machine learning [article]

Florian Steinke, Universität Des Saarlandes, Universität Des Saarlandes
Die vorgestellte Methode wird anhand verschiedener Beispiele aus der Robotik und der Computergraphik getestet.  ...  The novelty of our description is the broad, unifying view connecting kernel methods and linear system identification.  ...  A one pixel distance could be termed large, if it crosses an edge, and small otherwise.  ... 
doi:10.22028/d291-25950 fatcat:42nfjk6uanbo5eapt25cwhqtnu