"Previous researches have demonstrated that the framework of dictionary learning with sparse coding, in which signals are decomposed as linear combinations of a few atoms of a learned dictionary, is well adept to reconstruction issues. This framework has also been used for discrimination tasks such as image classification. To achieve better performances of classification, experts develop several methods to learn a discriminative dictionary in a supervised manner. However, another issue is that when the data become extremely large in scale, these methods will be no longer effective as they are all batch-oriented approaches. For this reason, we propose a novel online algorithm for discriminative dictionary learning, dubbed textbf{ODDL} in this paper. First, we introduce a linear classifier into the conventional dictionary learning formulation and derive a discriminative dictionary learning problem. Then, we exploit an online algorithm to solve the derived problem. Unlike the most existing approaches which update dictionary and classifier alternately via iteratively solving sub-problems, our approach directly explores them jointly. Meanwhile, it can largely shorten the runtime for training and is also particularly suitable for large-scale classification issues. To evaluate the performance of the proposed ODDL approach in image recognition, we conduct some experiments on three well-known benchmarks, and the experimental results demonstrate ODDL is fairly promising for image classification tasks."

"Single-particle imaging experiments of biomolecules at x-ray free-electron lasers (XFELs) require processing of hundreds of thousands (or more) of images that contain very few x-rays. Each low-flux image of the diffraction pattern is produced by a single, randomly oriented particle, such as a protein. We demonstrate the feasibility of collecting data at these extremes, averaging only 2.5 photons per frame, where it seems doubtful there could be information about the state of rotation, let alone the image contrast. This is accomplished with an expectation maximization algorithm that processes the low-flux data in aggregate, and without any prior knowledge of the object or its orientation. The versatility of the method promises, more generally, to redefine what measurement scenarios can provide useful signal in the high-noise regime."

"This paper proposes a novel adaptive algorithm to extract facial feature points automatically such as eyebrows corners, eyes corners, nostrils, nose tip, and mouth corners in frontal view faces, which is based on cumulative histogram approach by varying different threshold values. At first, the method adopts the Viola-Jones face detector to detect the location of face and also crops the face region in an image. From the concept of the human face structure, the six relevant regions such as right eyebrow, left eyebrow, right eye, left eye, nose, and mouth areas are cropped in a face image. Then the histogram of each cropped relevant region is computed and its cumulative histogram value is employed by varying different threshold values to create a new filtering image in an adaptive way. The connected component of interested area for each relevant filtering image is indicated our respective feature region. A simple linear search algorithm for eyebrows, eyes and mouth filtering images and contour algorithm for nose filtering image are applied to extract our desired corner points automatically. The method was tested on a large BioID frontal face database in different illuminations, expressions and lighting conditions and the experimental results have achieved average success rates of 95.27%."

"In this work we develop an algorithm for signal reconstruction from the magnitude of its Fourier transform in a situation where some (non-zero) parts of the sought signal are known. Although our method does not assume that the known part comprises the boundary of the sought signal, this is often the case in microscopy: a specimen is placed inside a known mask, which can be thought of as a known light source that surrounds the unknown signal. Therefore, in the past, several algorithms were suggested that solve the phase retrieval problem assuming known boundary values. Unlike our method, these methods do rely on the fact that the known part is on the boundary. Besides the reconstruction method we give an explanation of the phenomena observed in previous work: the reconstruction is much faster when there is more energy concentrated in the known part. Quite surprisingly, this can be explained using our previous results on phase retrieval with approximately known Fourier phase."

We consider the problem of building detectors for high-level concepts using only unsupervised feature learning. For example, we would like to understand if it is possible to learn a face detector using only unlabeled images downloaded from the internet. To answer this question, we trained a simple feature learning algorithm on a large dataset of images (10 million images, each image is 200x200). The simulation is performed on a cluster of 1000 machines with fast network hardware for one week. Extensive experimental results reveal surprising evidence that such high-level concepts can indeed be learned using only unlabeled data and a simple learning algorithm.

by memo(2009)

"Recently, the face recognizers based on linear representations have been shown to deliver state-of-the-art performance. In real-world applications, however, face images usually suffer from expressions, disguises and random occlusions. The problematic facial parts undermine the validity of the linear-subspace assumption and thus the recognition performance deteriorates significantly. In this work, we address the problem in a learning-inference-mixed fashion. By observing that the linear-subspace assumption is more reliable on certain face patches rather than on the holistic face, some Bayesian Patch Representations (BPRs) are randomly generated and interpreted according to the Bayes' theory. We then train an ensemble model over the patch-representations by minimizing the empirical risk w.r.t the "leave-one-out margins". The obtained model is termed Optimal Representation Ensemble (ORE), since it guarantees the optimality from the perspective of Empirical Risk Minimization. To handle the unknown patterns in test faces, a robust version of BPR is proposed by taking the non-face category into consideration. Equipped with the Robust-BPRs, the inference ability of ORE is increased dramatically and several record-breaking accuracies (99.9% on Yale-B and 99.5% on AR) and desirable efficiencies (below 20 ms per face in Matlab) are achieved. It also overwhelms other modular heuristics on the faces with random occlusions, extreme expressions and disguises. Furthermore, to accommodate immense BPRs sets, a boosting-like algorithm is also derived. The boosted model, a.k.a Boosted-ORE, obtains similar performance to its prototype. Besides the empirical superiorities, two desirable features of the proposed methods, namely, the training-determined model-selection and the data-weight-free boosting procedure, are also theoretically verified."

"This paper proposes a novel approach to image deblurring and digital zooming using sparse local models of image appearance. These models, where small image patches are represented as linear combinations of a few elements drawn from some large set (dictionary) of candidates, have proven well adapted to several image restoration tasks. A key to their success has been to learn dictionaries adapted to the reconstruction of small image patches. In contrast, recent works have proposed instead to learn dictionaries which are not only adapted to data reconstruction, but also tuned for a specific task. We introduce here such an approach to deblurring and digital zoom, using pairs of blurry/sharp (or low-/high-resolution) images for training, as well as an effective stochastic gradient algorithm for solving the corresponding optimization task. Although this learning problem is not convex, once the dictionaries have been learned, the sharp/high-resolution image can be recovered via convex optimization at test time. Experiments with synthetic and real data demonstrate the effectiveness of the proposed approach, leading to state-of-the-art performance for non-blind image deblurring and digital zoom."

バーコードだけでなく、表紙からでも認識が可能。Amazon.jpにも対応してほしい。

"In this paper we address the problem of understanding the success of algorithms that organize patches according to graph-based metrics. Algorithms that analyze patches extracted from images or time series have led to state-of-the art techniques for classification, denoising, and the study of nonlinear dynamics. The main contribution of this work is to provide a theoretical explanation for the above experimental observations. Our approach relies on a detailed analysis of the commute time metric on prototypical graph models that epitomize the geometry observed in general patch graphs.…"

"3D point clouds of natural environments relevant to geomorphology problems (rivers, cliffs...) often require to classify the data into elementary relevant classes. A typical example is the separation of riparian vegetation from soil in fluvial environments, the distinction between fresh surfaces and rockfall in cliff environments, or more generally the classification of surfaces according to their morphology (ripples, grain size...). Natural surfaces are very heterogeneous and their distinctive properties are seldom defined at a unique scale. We have thus defined a multi-scale measure of the point cloud dimensionality around each point. The dimensionality characterizes the local 3D organization of the point cloud and varies from being 1D (points set along a line) to really taking all 3D volume, at each scale. We present the technique and illustrate its efficiency in separating riparian vegetation from ground and classifying a mountain stream in vegetation, rock, gravel and water surface. The superiority of the multi-scale analysis in enhancing class separability and spatial resolution of the classification is also demonstrated. Large scenes can be classified on a commodity laptop in a reasonable time. The technique is robust to missing data and especially shadow zones. The classification is fast and accurate and can account for some degree of intra-class morphological variability such as different vegetation types. A probabilistic confidence in the classification result is given at each point allowing the user to remove the points for which the classification is uncertain. The process can be both fully automated but also fully customized by the user including a graphical definition of the classifiers if so desired. Although developed for fully 3D data, the method can be readily applied to 2.5D airborne LiDAR data."

A site that can examine the possibility of photoshopped images.

kinectでマルチタッチ・プロジェクション

監視カメラと鳥（カラス）の組み合わせが有機的。かわいくもあり、不気味。

"Sparse data models, where data is assumed to be well represented as a linear combination of a few elements from a dictionary, have gained considerable attention in recent years, and their use has led to state-of-the-art results in many signal and image processing tasks. It is now well understood that the choice of the sparsity regularization term is critical in the success of such models. Based on a codelength minimization interpretation of sparse coding, and using tools from universal coding theory, we propose a framework for designing sparsity regularization terms which have theoretical and practical advantages when compared to the more standard l0 or l1 ones. The presentation of the framework and theoretical foundations is complemented with examples that show its practical advantages in image denoising, zooming and classification."