Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “vision transformers”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 559 records · Page 31Linked to original sources
PubMed1994

Spectral sharpening: sensor transformations for improved color constancy.

We develop sensor transformations, collectively called spectral sharpening, that convert a given set of sensor sensitivity functions into a new set that will improve the performance of any color-constancy algorithm that is based on an independent adjustment of the sensor response channels. Independent adjustment of multiplicative coefficients corresponds to the application of a diagonal-matrix transform (DMT) to the sensor response vector and is a common feature of many theories of color constancy. Land's retinex and von Kries adaptation in particular. We set forth three techniques for spectral sharpening. Sensor-based sharpening focuses on the production of new sensors as linear combinations of the given ones such that each new sensor has its spectral sensitivity concentrated as much as possible within a narrow band of wavelengths. Data-based sharpening, on the other hand, extracts new sensors by optimizing the ability of a DMT to account for a given illumination change by examining the sensor response vectors obtained from a set of surfaces under two different illuminants. Finally in perfect sharpening we demonstrate that, if illumination and surface reflectance are described by two- and three-parameter finite-dimensional models, there exists a unique optimal sharpening transform. All three sharpening methods yield similar results. When sharpened cone sensitivities are used as sensors, a DMT models illumination change extremely well. We present simulation results suggesting that in general nondiagonal transforms can do only marginally better. Our sharpening results correlate well with the psychophysical evidence of spectral sharpening in the human visual system.

Algorithms↗
PubMed2002

Integrated imaging in peripheral nerve lesions in type 1 neurofibromatosis.

PURPOSE: Our aim was to evaluate the role of Ultrasonography (US) and Magnetic Resonance (MR) in the staging and follow-up of peripheral nerve lesions in type 1 Neurofibromatosis (NF1). MATERIAL AND METHODS: We evaluated forty-six NF1 patients (16 males and 30 females, age range 12-65 years, mean age 34 years) affected by 51 soft tissue lesions, clinically diagnosed as subcutaneous (18 cases) and plexiform (33 cases) neurofibromas. The lesions were studied to identify site, size, extension, relationship with surrounding structures, and any features indicating malignant transformation. All patients underwent at least one US examination performed with a 7.5 - 10 MHz linear multi-frequency probe. All subcutaneous neurofibromas larger than 2 cm, all superficial or deep plexiform neurofibromas not clearly defined at US, and all large plexiform neurofibromas with massive involvement of surrounding tissues were studied by MR using 0.2 T magnet (Artoscan) or 1.5 T magnet (Magnetom Vision Plus Siemens). Fifteen patients underwent surgery; the remaining cases (31) were followed up according to the National Institutes of Health (NIH) protocol. RESULTS: The subcutaneous neurofibromas smaller than 2 cm showed homogeneous hypoechoic echotexture (10 cases), whereas lesions equal to or larger than 2 cm showed a characteristic target pattern (peripheral hypoechoic rim with hyperechoic core). All lesions larger than 2 cm studied by MRI displayed the target pattern in T2-weighted sequences, with peripheral signal hyperintensity and central signal hypointensity. At US, the 33 plexiform neurofibromas had an irregular shape with undefined borders; the patterns were classified as follows: 1) superficial plate-like lesions (14 cases); 2) superficial lesions with deep digitations (8 cases); 3) deep multi-lobulated lesions (5 cases); 4) extensive lesions massively involving soft tissues (6 cases). MRI was performed in 11 cases, showing an isointense or moderately hyperintense signal on T1-weighted sequences, high signal intensity on T2-weighted sequences, and, after intravenous gadolinium administration, variable enhancement on T1-weighted sequences. CONCLUSIONS: The results obtained in subcutaneous neurofibromas indicate a clear relationship between lesion size and US-MR imaging, with similar local staging. Since MRI failed to provide additional information, US may be considered the first-choice, and sometimes decisive, examination in these cases. As regards plexiform neurofibromas, US only proved useful in staging superficial lesions and partly deep multi-lobulated lesions as it accurately depicted lesion size, but not local extension; MRI is therefore useful in the preoperative staging of lesions. In extensive lesions, US proved unable to provide accurate and complete local staging of the lesions, so that MRI should be systematically used in these cases.

Adolescent↗
PubMed1997

Visual space distortion.

We are surrounded by surfaces that we perceive by visual means. Understanding the basic principles behind this perceptual process is a central theme in visual psychology, psychophysics, and computational vision. In many of the computational models employed in the past, it has been assumed that a metric representation of physical space can be derived by visual means. Psychophysical experiments, as well as computational considerations, can convince us that the perception of space and shape has a much more complicated nature, and that only a distored version of actual, physical space can be computed. This paper develops a computational geometric model that explains why such distortion might take place. The basic idea is that, both in stereo and motion, we perceive the world from multiple views. Given the rigid transformation between the views and the properties of the image correspondence, the depth of the scene can be obtained. Even a slight error in the rigid transformation parameters causes distortion of the computed depth of the scene. The unified framework introduced here describes this distortion in computational terms. We characterize the space of distortions by its level sets, that is, we characterize the systematic distortion via a family of iso-distortion surfaces which describes the locus over which depths are distorted by some multiplicative factor. Given that humans' estimation of egomotion or estimation of the extrinsic parameters of the stereo apparatus is likely to be imprecise, the framework is used to explain a number of psychophysical experiments on the perception of depth from motion or stereo.

Cybernetics↗
PubMed2004

Mirror vision: perceived size and perceived distance of virtual images.

We investigated spatial perception of virtual images that were produced by convex and plane mirrors. In Experiment 1, 36 subjects reproduced both the perceived size and the perceived distance of virtual images for five targets that had been placed at a real distance of 10 or 20 m. In Experiment 2, 30 subjects verbally judged both the perceived size and the perceived distance of virtual images for five targets that were placed at each of five real distances of 2.5-45 m. In both experiments, the subjects received objective-size and objective-distance instructions. The results were that (1) size constancy was attained for a distance of up to 45 m, (2) distance was readily discriminated within this distance range, although virtual images produced by the mirror of strong curvature were judged to be farther away than those produced by the mirrors of less curvature, and (3) the ratio of perceived size to perceived distance was described as a power function of visual angle, and the ratio for the convex mirror was larger than that for the plane mirror. We compared the taking-into-account model and the direct perception model on the basis of a correlation analysis for proximal, virtual, and real levels of the stimuli. The taking-into-account model, which assumes that visual angle is transformed into perceived size by taking perceived distance into account, was supported by an analysis for the proximal level of stimuli. The direct perception model, which assumes that there is no inferential process between perceived size and perceived distance, was partially supported by an analysis for the distal level of the stimuli.

Adult↗
PubMed2002

[Adaptation of a dementia test for use in visually handicapped persons exemplified by SIDAM (Structured Interview for the Diagnosis of Alzheimer disease, Multi-Infarct Dementia and Dementias of other Etiology)].

In epidemiological field studies on the prevalence and incidence of dementia, the problem of cognitive testing of visually impaired individuals is rarely discussed. In the Leipzig Longitudinal Study of the Aged (LEILA 75+) a version of the SIDAM (SIDAM-Structured Interview for the Diagnosis of dementia of the Alzheimer type, Multi-infarct dementia and dementias of other etiology) for the visually impaired was employed from which all items requiring image processing had been omitted (SIDAM-blind). In order to be able to interpret the test results, the final scores for the full SIDAM were estimated by linear transformation of the scores in the blind-version. The method of linear transformation is based on certain theoretical assumptions which are examined in this paper. Linear transformation of scores has proved to be a valid procedure only for individuals with high cognitive performances. Thus, an evaluation of estimated scores based on the norms for the original SIDAM has limitations. Consequently, age-specific norms for the SIDAM-blind are presented. Sensitivities and specificities for different cut-off points used to discriminate between demented and non-demented vision-impaired individuals are given.

Aged↗
PubMedDate not supplied

Chemical machine vision: automated extraction of chemical metadata from raster images.

We present a novel application of machine vision methods for the identification of chemical composition diagrams from two-dimensional digital raster images. The method is based on the use of Gabor wavelets and an energy function to derive feature vectors from digital images. These are used for training and classification purposes using a Kohonen network for classification with the Euclidean distance norm. We compare this method with previous approaches to transforming such images to a molecular connection table, which are designed to achieve complete atom connection table fidelity but at the expense of requiring human interaction. The present texture-based approach is complementary in attempting to recognize higher order features such as the presence of a chemical representation in the original raster image. This information can be used for providing chemical metadata descriptors of the original image as part of a robot-based Internet resource discovery tool.

Data Display↗
PubMed1997

Visual-motor transformations required for accurate and kinematically correct saccades.

The goal of this study was to identify and model the three-dimensional (3-D) geometric transformations required for accurate saccades to distant visual targets from arbitrary initial eye positions. In abstract 2-D models, target displacement in space, retinal error (RE), and saccade vectors are trivially interchangeable. However, in real 3-D space, RE is a nontrivial function of objective target displacement and 3-D eye position. To determine the physiological implications of this, a visuomotor "lookup table" was modeled by mapping the horizontal/vertical components of RE onto the corresponding vector components of eye displacement in Listing's plane. This provided the motor error (ME) command for a 3-D displacement-feedback loop. The output of this loop controlled an oculomotor plant that mechanically implemented the position-dependent saccade axis tilts required for Listing's law. This model correctly maintained Listing's law but was unable to correct torsional position deviations from Listing' s plane. Moreover, the model also generated systematic errors in saccade direction (as a function of eye position components orthogonal to RE), predicting errors in final gaze direction of up to 25 degrees in the oculomotor range. Plant modifications could not solve these problems, because the intrisic oculomotor input-output geometry forced a fixed visuomotor mapping to choose between either accuracy or Listing's law. This was reflected internally by a sensorimotor divergence between input-defined visual displacement signals (inherently 2-D and defined in reference to the eye) and output-defined motor displacement signals (inherently 3-D and defined in reference to the head). These problems were solved by rotating RE by estimated 3-D eye position (i.e., a reference frame transformation), inputting the result into a 2-D-to-3-D "Listing's law operator," and then finally subtracting initial 3-D eye position to yield the correct ME. This model was accurate and upheld Listing's law from all initial positions. Moreover, it suggested specific experiments to invasively distinguish visual and motor displacement codes, predicting a systematic position dependence in the directional tuning of RE versus a fixed-vector tuning in ME. We conclude that visual and motor displacement spaces are geometrically distinct such that a fixed visual-motor mapping will produce systematic and measurable behavioral errors. To avoid these errors, the brain would need to implement both a 3-D position-dependent reference frame transformation and nontrivial 2-D-to-3-D transformation. Furthermore, our simulations provide new experimental paradigms to invasively identify the physiological progression of these spatial transformations by reexamining the position-dependent geometry of displacement code directions in the superior colliculus, cerebellum, and various cortical visuomotor areas.

Animals↗
PubMed1999

Learning to design synergetic computers with an extended symmetric diffusion network.

This article proposes an extended symmetric diffusion network that is applied to the design of synergetic computers. The state of a synergetic computer is translated to that of order parameters whose dynamics is described by a stochastic differential equation. The order parameter converges to the Boltzmann distribution, under some condition on the drift term, derived by the Fokker-Planck equation. The network can learn the dynamics of the order parameters from a nonlinear potential. This property is necessary to design the coefficient values of the synergetic computer. We propose a searching function for the image processing executed by the synergetic computer. It is shown that the image processing with the searching function is superior to the usual image-associative function of synergetic computation. The proposed network can be related, as a special case, to the discrete-state Boltzmann machine by some transformation. Finally, the extended symmetric diffusion network is applied to the estimation problem of an entire density function, as well as the proposed searching function for the image processing.

Artificial Intelligence↗
PubMed2006

Malignant prolactinoma: case report and review of the literature.

Pituitary carcinomas are extremely rare. In general, the initial clinical, biochemical, and histological characteristics are of minimal utility in distinguishing benign adenomas from pituitary carcinomas. We describe a 63-year-old woman with a macroprolactinoma, who presented with diplopia and blurred vision. This unusual initial presentation and the subsequent aggressive clinical course, with diffuse local and distant intramedullary metastases, prompted us in retrospect to make a detailed analysis of the therapeutic interventions and histology. In addition, we reviewed all available literature on published cases of malignant prolactinoma and detailed their epidemiological, clinical, and histopathological characteristics. In brief, it is postulated that pituitary carcinomas arise from the transformation of initially large, but benign, adenomas. Unusual and/or atypical clinical manifestations appear to occur more frequently. In vivo, the development of dopamine agonist resistance in invasive macroprolactinoma is indicative of malignancy and should prompt the clinician to perform a biopsy of the tumor. For pituitary tumors that exhibit high mitotic activity, increased Ki-67 and/or p53 immunoreactivity, it may be useful to denote these tumors as 'atypical' prolactinomas to raise the possibility of future malignant development.

Adolescent↗
PubMed1995

Learning a visuomotor transformation in a local area of work space produces directional biases in other areas.

1. The dependence of directional biases in reaching movements on the initial position of the hand was studied in normal human subjects moving their unseen hand on a horizontal digitizing tablet to visual targets displayed on a vertical computer screen. 2. When initial hand positions were to the right of midline, movements were systematically biased clockwise. Biases were counterclockwise for starting points to the left. Biases were unaffected by the screen location of the starting and target positions. 3. Vision of the hand in relation to the target before movement, as well as practice with vision of the cursor during the movement, temporarily eliminated these biases. The spatial organization of the biases suggests that, without vision of the limb, the nervous system underestimates the distance of the hand from an axis or plane that includes its most common operating location. 4. To test the hypothesis that such an underestimate might represent an adaptation to a local area of work space or range effect, subjects were trained to reach accurately from right or left positions. After training, movements initiated from other locations, including ones that were previously error free, showed new biases that again represented underestimates of the distance of the initial hand position from the new trained location. 5. We conclude that hand path planning is dependent on learned representations of the location of the hand in the work space.

Adult↗
PubMed1994

[Phylogenetic vision of bile acids].

Bile acids are the most important solutes of bile: they are essential in cholesterol degradation, solubilization and excretion; they are determinants of bile flow and secretion; and their role is crucial in the intestinal absorption of lipids and lipid soluble vitamins. In amphibia and in cartilaginous fish, the 27C cholestane molecule is hydroxylated to alcohols. In birds, the terminal 27C-OH group is oxydated to cholestanoic acids. In vertebrates of a more recent evolutionary origin, the lateral chain is shortened to 24C and oxydated to cholestanoic acids. Further transformations include chemical changes in the cholestane skeleton and in the lateral chain (hydroxylations, dehydroxylations, epimerization, etc). In the intestinal lumen, the saprophytic flora provides enzymes catalysing new changes that originate "secondary" bile acids. During entero-hepatic circulation, another variety of bile acids appear, commonly termed "tertiary" bile acids. A recent study of Lee R Hagey characterized bile acid composition of over 600 species of vertebrates, showing that bile acid composition of bile has been the subject of an interesting evolutionary phenomenon and that it is a chemical marker of biodiversity in vertebrates.

Animals↗
PubMed1986

Translation of sensory signals into commands for control of saccadic eye movements: role of primate superior colliculus.

Afferent signals that guide orienting movements converge in the deeper layers of the SC in a wide variety of animals. The sensory cells are arranged topographically according to their receptive-field locations and, thereby, form maps of sensory space. Maps of visual, somatosensory, and/or auditory space have been obtained in the iguana, mouse, hamster, barn owl, chinchilla, cat, and monkey. The deeper layers of the SC also contain neurons involved in the generation of movements of the eyes, head, vibrissae, and pinnae. Thus the SC, a site containing multiple sensory maps and perhaps multiple motor maps, has been selected by many investigators as a structure for investigating the problem of sensorimotor integration. In the mammalian nervous system, emphasized in this review, much remains to be learned about the structure, organization, and function of the SC. While anatomical studies continue to add to the knowledge of the sources of afferent projections, their pattern of laminar termination, and the source and destination of efferent projections, relatively little is known about the intrinsic organization of the colliculus, especially the deeper layers. Recently, electrophysiological studies have moved from an emphasis on the sensory and motor properties of collicular neurons to an examination of the maps of auditory and somatosensory space and the correspondence of these maps. In the future, major efforts aimed at identifying the functional properties of cells that project to the SC from diverse brain regions as well as the functional properties that project to the various structures receiving input from the colliculus are needed. A combination of anatomical and electrophysiological methods is required to describe the signal transforms that occur between the SC and motor areas (such as the paramedian pontine reticular formation) closer to the final common pathway. Conceptual and empirical work is needed to develop and test models of how the dynamic visual and auditory maps found in the primate SC are generated. In general, new and/or improved models of the role of the SC in sensorimotor integration are needed as guides for future research. A point of view emphasized here is that it may be fruitful to examine the function of the SC from a motor perspective. The nature of the motor command imposes constraints on the configuration of signals that can initiate movements and thereby determines the required transformation of sensory signals.

Acoustic Stimulation↗
PubMed2003

[Calculating the mean visual acuity and the change in visual acuity with a decimal acuity chart].

The decimal visual acuity chart is not easy to use for statistical analysis and requires a transformation into LogMAR units (decimal logarithm of the Minimum Angle Resolution). In contrast to the decimal chart, the logarithmic chart has an arithmetic progression and a constant interval between lines. The LogMAR chart makes statistical analysis of visual acuity easy. Change in visual acuity is calculated directly by subtracting LogMAR data, while the average visual acuity is obtained with the arithmetic mean value of the LogMAR data. The mean acuity expressed in LogMAR units can be transformed into a decimal chart for a more comprehensive result. To calculate the average visual acuity directly from the decimal data, the geometric mean value must be used instead of the arithmetic mean value.

Calibration↗
PubMed1981

Reorganization of perception: a Konorskian interpretation of the Innsbruck experiments.

This paper discusses the close fit of Konorski’s model to the data of Kohler’s Innsbruck experiments and also to more recent neurophysiological studies of environmental surgery or sensory disturbances. The dramatic Innsbruck research using special goggles or prismatic glasses that reversed or distorted the visual field confirmed and extended Stratton's earlier use of the substitution method. The subjects had at first insuperable difficulties in their locomotor behavior if it was guided by vision. After practice guided by certain rules, they were able to perform and eventually to "see" correctly in the strange environment. These studies demonstrate the plasticity of the nervous system, even in the adult. The problem for the field is to determine how these new associations, antagonistic to long established conditioned reflexes (CRs), become functional. The particular questions arise: what rules impede, what rules facilitate the transformation of CRs, and what physiological mechanism is responsible for the changed relations between the organism and the external world? Konorski’s remarkable theory is that no transformations of connections occurs in the reversal training. Instead, what actually happens is the formation of a connection between the original instrumental CR and non-reinforcement, and actualization of excitatory connections between the CS units and the kinesthetic units of the antagonistic movement. If correct, this motor act marks the termination of the eliciting drive and is reinforced. The original CR is readily re-established by restitution of its reinforcement.

Adaptation, Physiological↗
PubMed1994

Correspondence-based and energy-based detection of second-order motion in human vision.

Motion in the retinal image may occur either in the form of spatiotemporal variations in luminance (first-order motion) or as spatiotemporal variations in characteristics derived from luminance, such as contrast (second-order motion). Second-order motion patterns were employed in an attempt to establish the principles used for the detection of image motion in the human visual system. In principle, one can detect motion at a high level of visual analysis by identifying features of the image and tracking their positions (correspondence-based detection) or at a low level by analysis of spatiotemporal luminance variations without reference to features (intensity-based detection). Prevailing models favor the latter approach, which has been adapted to account for the visibility of second-order motion by postulation of a stage of rectification that precedes motion energy detection [J. Opt. Soc. Am. A 5, 1986 (1988)]. In two experiments it is shown that second-order motion is indeed detected normally by use of the strategy of transformation plus energy detection but that detection can also be achieved by use of the feature-correspondence strategy when the intensity strategy fails. In the first experiment, a stimulus is employed in which opposite directions of motion perception are predicted by the two strategies. It is shown that normally the direction associated with motion energy in the rectified image is seen but that the direction associated with feature motion is seen when the energy system is disabled by the use of an interstimulus interval.(ABSTRACT TRUNCATED AT 250 WORDS)

Contrast Sensitivity↗
PubMed2007

Direct curvature scale space: theory and corner detection.

The Curvature Scale Space (CSS) technique is considered to be a modern tool in image processing and computer vision. Direct Curvature Scale Space (DCSS) is defined as the CSS that results from convolving the curvature of a planar curve with a Gaussian kernel directly. In this paper we present a theoretical analysis of DCSS in detecting corners on planar curves. The scale space behavior of isolated single and double corner models is investigated and a number of model properties are specified which enable us to transform a DCSS image into a tree organization and, so that corners can be detected in a multiscale sense. To overcome the sensitivity of DCSS to noise, a hybrid strategy to apply CSS and DCSS is suggested.

Algorithms↗
PubMed1998

Focused attention and the detectability of weak gustatory stimuli. Empirical measurement and computer simulations.

Attentional processes can modulate the detectability of weak stimuli; for example, the detectability of visual or auditory signals can depend on whether attention is allocated to the appropriate spatial location (vision) or acoustic frequency (hearing). Earlier attempts in the first author's laboratory to find analogous effects of focused attention on the detectability of taste stimuli were equivocal, in part it seems because human gustatory sensitivity can fluctuate substantially over time, a serious problem when using procedures that track sensitivity (d') to a constant stimulus concentration. To circumvent this problem, we adopted an adaptive psychophysical procedure, the transformed up-down method, using a 3-down/1-up rule to determine how the threshold to detect weak concentrations of sucrose and citric acid depended on whether the stimulus presented in a given two-alternative, forced-choice trial was expected or unexpected. The results showed threshold sensitivity to be slightly but consistently poorer when the test stimulus was unexpected (e.g., sucrose presented when citric acid was expected) than it was when the test stimulus was expected (e.g., sucrose presented when sucrose was expected). In this attentional paradigm, the unexpected stimulus must perforce be presented on only a small fraction of the trials. In selecting a procedure, we chose a 3-down/1-up adaptive rule rather than the more popular 2-down/1-up rule, a choice that turned out to be in line with results of Monte Carlo computer simulations. These simulations suggest that across a wide range of conditions (starting stimulus concentrations, step sizes), the variability in threshold measurements can be smaller with a 3-down/1-up rule than with a 2-down/1-up rule, even when the total number of trials is the same and not very great.

Attention↗
PubMed2006

Ecology for transformation.

Ecology has a key role in our understanding of the benefits that humans obtain from ecosystems (i.e. ecosystem services). Ecology can also contribute to developing environmentally sound technologies, markets for ecosystem services and approaches to decision-making that account for the changing relationship between humans and ecosystems. These contributions involve basic ecological research on, for example, the resilience of ecosystem services or relationships of ecosystem change to natural disasters. Much of the necessary work involves interdisciplinary collaboration among ecologists, social scientists and decision makers. As we discuss here, ecology should help formulate positive, plausible visions for relationships of society and ecosystems that can potentially sustain ecosystem services for long periods of time.

Biodiversity↗