111 |
Contributions à l'apprentissage grande échelle pour la classification d'imagesAkata, Zeynep 06 January 2014 (has links) (PDF)
La construction d'algorithmes classifiant des images à grande échelle est devenue une tache essentielle du fait de la difficulté d'effectuer des recherches dans les immenses collections de données visuelles inetiquetées présentes sur Internet. Nous visons à classifier des images en fonction de leur contenu pour simplifier la gestion de telles bases de données. La classification d'images à grande échelle est un problème complèxe, de par l'importance de la taille des ensembles de données, tant en nombre d'images qu'en nombre de classes. Certaines de ces classes sont dites "fine-grained" (sémantiquement proches les unes des autres) et peuvent même ne contenir aucun représentant étiqueté. Dans cette thèse, nous utilisons des représentations état de l'art d'images et nous concentrons sur des méthodes d'apprentissage efficaces. Nos contributions sont (1) un banc d'essai d'algorithmes d'apprentissage pour la classification à grande échelle et (2) un nouvel algorithme basé sur l'incorporation d'étiquettes pour apprendre sur des données peu abondantes. En premier lieu, nous introduisons un banc d'essai d'algorithmes d'apprentissage pour la classification à grande échelle, dans le cadre entièrement supervisé. Il compare plusieurs fonctions objectifs pour apprendre des classifieurs linéaires, tels que "un contre tous", "multiclasse", "ranking", "ranking pondéré moyen" par descente de gradient stochastique. Ce banc d'essai se conclut en un ensemble de recommandations pour la classification à grande échelle. Avec une simple repondération des données, la stratégie "un contre tous" donne des performances meilleures que toutes les autres. Par ailleurs, en apprentissage en ligne, un pas d'apprentissage assez petit s'avère suffisant pour obtenir des résultats au niveau de l'état de l'art. Enfin, l'arrêt anticipé de la descente de gradient stochastique introduit une régularisation qui améliore la vitesse d'entraînement ainsi que la capacité de régularisation. Deuxièmement, face à des milliers de classes, il est parfois difficile de rassembler suffisamment de données d'entraînement pour chacune des classes. En particulier, certaines classes peuvent être entièrement dénuées d'exemples. En conséquence, nous proposons un nouvel algorithme adapté à ce scénario d'apprentissage dit "zero-shot". notre algorithme utilise des données parallèles, comme les attributs, pour incorporer les classes dans un espace euclidien. Nous introduisons par ailleurs une fonction pour mesurer la compatibilité entre image et étiquette. Les paramètres de cette fonction sont appris en utilisant un objectif de type "ranking". Notre algorithme dépasse l'état de l'art pour l'apprentissage "zero-shot", et fait preuve d'une grande flexibilité en permettant d'incorporer d'autres sources d'information parallèle, comme des hiérarchies. Il permet en outre une transition sans heurt du cas "zero-shot" au cas où peu d'exemples sont disponibles.
|
112 |
X-ray computed tomography reconstruction on non-standard trajectories for robotized inspection / Reconstruction 3D en tomographie X sur trajectoires non-standardsBanjak, Hussein 10 November 2016 (has links)
La tomographie par rayons X ou CT pour "Computed Tomography" est un outil puissant pour caractériser et localiser les défauts internes et pour vérifier la conformité géométrique d’un objet. Contrairement au cas des applications médicales, l’objet inspecté en Contrôle Non Destructif (CND) peut être très grand et composé de matériaux de haute atténuation, auquel cas l’utilisation d’une trajectoire circulaire pour l’inspection est impossible à cause de contraintes dans l’espace. Pour cette raison, l’utilisation de bras robotisés est l’une des nouvelles tendances reconnues dans la CT, car elle autorise plus de flexibilité dans la trajectoire d’acquisition et permet donc la reconstruction 3D de régions difficilement accessibles dont la reconstruction ne pourrait pas être assurée par des systèmes de tomographie industriels classiques. Une cellule de tomographie X robotisée a été installée au CEA. La plateforme se compose de deux bras robotiques pour positionner et déplacer la source et le détecteur en vis-à-vis. Parmi les nouveaux défis posés par la tomographie robotisée, nous nous concentrons ici plus particulièrement sur la limitation de l’ouverture angulaire imposée par la configuration en raison des contraintes importantes sur le mouvement mécanique de la plateforme. Le deuxième défi majeur est la troncation des projections qui se produit lorsque l’objet est trop grand par rapport au détecteur. L’objectif principal de ce travail consiste à adapter et à optimiser des méthodes de reconstruction CT pour des trajectoires non standard. Nous étudions à la fois des algorithmes de reconstruction analytiques et itératifs. Avant d’effectuer des inspections robotiques réelles, nous comptons sur des simulations numériques pour évaluer les performances des algorithmes de reconstruction sur des configurations d’acquisition de données. Pour ce faire, nous utilisons CIVA, qui est un outil de simulation pour le CND développé au CEA et qui est capable de simuler des données de projections réalistes correspondant à des configurations d’acquisition définies par l’utilisateur. / X-ray computed tomography (CT) is a powerful tool to characterize or localize inner flaws and to verify the geometric conformity of an object. In contrast to medical applications, the scanned object in non-destructive testing (NDT) might be very large and composed of high-attenuation materials and consequently the use of a standard circular trajectory for data acquisition would be impossible due to constraints in space. For this reason, the use of robotic arms is one of the acknowledged new trends in NDT since it allows more flexibility in acquisition trajectories and therefore could be used for 3D reconstruction of hardly accessible regions that might be a major limitation of classical CT systems. A robotic X-ray inspection platform has been installed at CEA LIST. The considered system integrates two robots that move the X-ray generator and detector. Among the new challenges brought by robotic CT, we focus in this thesis more particularly on the limited access viewpoint imposed by the setup where important constraints control the mechanical motion of the platform. The second major challenge is the truncation of projections that occur when only a field-of-view (FOV) of the object is viewed by the detector. Before performing real robotic inspections, we highly rely on CT simulations to evaluate the capability of the reconstruction algorithm corresponding to a defined scanning trajectory and data acquisition configuration. For this purpose, we use CIVA which is an advanced NDT simulation platform developed at CEA and that can provide a realistic model for radiographic acquisitions and is capable of simulating the projection data corresponding to a specific CT scene defined by the user. Thus, the main objective of this thesis is to develop analytical and iterative reconstruction algorithms adapted to nonstandard trajectories and to integrate these algorithms in CIVA software as plugins of reconstruction.
|
113 |
Ultracold atoms in trapsSala, Simon Johannes 08 April 2016 (has links)
Diese Dissertation widmet sich der theoretischen Beschreibung ultrakalter Atome in einem optischen Einschluss. Das Hauptaugenmerk liegt hierbei auf inelastischen Resonanzen, die durch die Kopplung von Schwerpunkts- und Relativbewegung durch Anharmonizitäten im externen Potenzial Zustande kommen, der Entwicklung einer Methode zur theoretischen Beschreibung von ultrakalten Wenigteilchensystemen in einem vielseitigen Einschlusspotenzial und der Quantensimulation von Attosekundenphysik mit ultrakalten Atomen. / This thesis aims for a theoretical description of ultracold trapped atoms. The main focus are resonance phenomena due to the coupling of center-of-mass and relative motion, the development of a theoretical approach to treat ultracold few-body systems in versatile trap potentials, and the quantum simulation of attosecond physics with ultracold atoms.
|
114 |
Gender, Myth, and Warfare: A Cross-Cultural Analysis of Women WarriorsUnknown Date (has links)
A combination of cross-cultural and symbolic methodologies suggests that
women warriors occur in societies where there is both an emphasis on the sacred
feminine that allows women greater access to positions of power and authority (as per
Peggy Sanday) and where marital residency rules permit female fighters (following
David B. Adams´s theory on women warriors). While neither theory can stand alone in
explaining the existence of women warriors, when combined both theories give a solid
picture of societies that allow for female combatants. In this paper I propose that by
combining Sanday’s work on female power and Adams’s work on women warriors we
can come to a better understanding about just what makes the cultures that allow for
women’s participation in warfare unique, and perhaps what characteristics must be in
place in order for a culture to have women warriors. / Includes bibliography. / Thesis (M.A.)--Florida Atlantic University, 2016. / FAU Electronic Theses and Dissertations Collection
|
115 |
Production and interaction of photons using atomic polaritons and Rydberg interactions / Production et interaction de photons en utilisant des polaritons atomiques et des interactions de RydbergBimbard, Erwan 01 December 2014 (has links)
Produire et faire interagir entre eux des photons optiques de façon contrôlée sont deux conditions nécessaires au développement de communications quantiques à longue distance, et plus généralement au traitement quantique d’information codée sur des photons. Cette thèse présente une étude expérimentale de solutions possibles a ces deux problèmes, en utilisant la conversion des photons en excitations collectives (polaritons) dans un nuage d’atomes froids, placé dans le mode d’une cavité optique de faible finesse (~100). Dans un premier temps, des polaritons entre états atomiques fondamentaux sont utilisés pour « mettre en mémoire » une excitation unique dans le nuage. Celle-ci est ensuite convertie efficacement en un photon unique, dont le champ est analysé par tomographie homodyne. La fonction de Wigner de l’état à un photon est reconstruite a partir des données expérimentales, et présente des valeurs négatives, démontrant que les degrés de liberté de ce photon (mode spatio-temporel et état quantique) sont complètement contrôlés. Dans un second temps, les photons sont couplés à des polaritons impliquant des états de Rydberg. Les fortes interactions dipolaires entre ces derniers se traduisent par des non-linéarités optiques dispersives très importantes, qui sont caractérisées dans un régime d’excitation classique. Ces non-linéarités peuvent être amplifiées jusqu’à ce qu’un seul photon suffise à modifier totalement la réponse du système, permettant en principe de générer des interactions effectives entre photons. / Controllably producing optical photons and making them interact are two key requirements for the development of long-distance quantum communications, and more generally for photonic quantum information processing. This thesis presents experimental studies on possible solutions to these two problems, using the conversion of the photons into collective excitations (polaritons) in a cold atomic cloud, inside the mode of a low-finesse optical cavity (~100). Firstly, ground-state polaritons are used to store a single excitation in the cloud memory. This polariton is then efficiently converted into a single photon, whose field is characterized via homodyne tomography. The single photon state’s Wigner function is reconstructed from the experimental data and exhibits negative values, demonstrating that the photon’s degrees of freedom (spatio-temporal mode and quantum state) are well controlled. Secondly, photons can be coupled to polaritons involving Rydberg states. The strong dipolar interactions between these give rise to very strong optical dispersive nonlinearities, that are characterized in a classical excitation regime. These nonlinearities can be amplified until a single photon is enough to modify the entire system’s response, allowing in principle for the generation of effective photon-photon interactions.
|
116 |
Industrialização e desenvolvimento do sudoeste do Paraná / Industrialization and development of the Paraná Southwest.Flores, Edson Luiz 14 July 2009 (has links)
Made available in DSpace on 2017-07-10T17:30:55Z (GMT). No. of bitstreams: 1
CAPITULO I.pdf: 3435123 bytes, checksum: 3a217406f96215893a8f341510f2421a (MD5)
Previous issue date: 2009-07-14 / How wrote Marx and Engels (1998), the production is very important to man, more
than this, it`s the responsible by own humanization; to measure that the men, to
opposite of the animals that live by nature (offer of foods, resisting by climatic
intemperance etc), get to produce the instrument of survival (foods, clothes, home
etc). However, we verified that in the human sciences, specifically at the Geography,
in the last epochs has been given little attention to the production, mainly to
industrialization. In this study, we realized a analysis about the industry`s importance
at the Southwest of the Parana`s development, region of new population occupation,
happened mainly after the 1940`s. By the realization of that study, proceeded to light
from the historic materialism, method which to consist, thick manner, in to analyses
the formations and transformations occurred at the society from active production
and from work`s connections intrinsic to its. The theoretical reference Marxism
mainly the Karl Marx, Friedrich Engels, Vladimir Lenin, Karl Kautsky, Ignácio Rangel,
Armem Mamigonian`s studies, among others abetted us to that realized a detailed
research, interviewing owns and executives of industrial unities from region, salaried
workers etc, looking at to get information and dies about the thematic studied. In this
discourse, we analyzed: 1) the importance of the industry (mainly the wood`s branch)
to the population of this region from Parana; 2) we verified the conditions that the
social formation (marked by presence of few owns of earth, craftsmen and
merchants) offered to organize a few merchant production; 3) we saw the importance
that had the financial capital, as modernizing the agricultural as financing the
industrial production (mainly from the second half at 1970`s); 4) we saw the
importance that the abetment politics to industrialization, the offer of raw materials,
and of work`s power, have been to attract industrial unities to this region; and, 5) we
verified the increase of income advanced by industry, that besides, has been
contributed, somehow, to a geral advancement of the life`s conditions from
Southwest of the Parana`s population. / Como escreveram Marx e Engels (1998), a produção é de suma importância ao
homem, mais do que isso, ela é responsável pela própria humanização; à medida
que os homens, ao contrário dos animais que vivem à mercê da natureza (da oferta
de alimentos, resistindo às intempéries climáticas etc.), conseguem produzir os
meios de sobrevivência dos quais necessitam (alimentos, vestuário, residências
etc.). No entanto, verificamos que nas ciências humanas, especificamente na
Geografia, nos últimos tempos tem-se dado pouca atenção à produção,
principalmente à industrialização. Nessa pesquisa, realizamos uma análise da
importância da indústria no desenvolvimento do Sudoeste do Paraná, região de
ocupação populacional recente, ocorrida principalmente após a década de 1940.
Para a realização de tal estudo, procedemos à luz do materialismo histórico, método
que consiste, grosso modo, em analisar as formações e transformações ocorridas na
sociedade a partir da organização produtiva e das relações de trabalho inerentes a
ela. O referencial teórico marxista principalmente estudos de Karl Marx, Friedrich
Engels, Vladimir Lenin, Karl Kautsky, Ignácio Rangel, Armem Mamigonian, entre
outros nos apoiou para que realizássemos uma pesquisa minuciosa, entrevistando
proprietários e/ou administradores de unidades industriais da região, trabalhadores
assalariados etc., visando obter informações e dados acerca da temática estudada.
Nessa dissertação, analisamos: 1) a importância da indústria (principalmente o ramo
da madeira) para o povoamento dessa região do Paraná; 2) verificamos as
condições que a formação social (marcada pela presença de pequenos proprietários
de terra, artesãos e comerciantes) ofereceu ao constituir uma pequena produção
mercantil; 3) observamos a importância que o capital financeiro teve, tanto
modernizando a agricultura como financiando a produção industrial (principalmente a
partir da segunda metade da década de 1970); 4) observamos a importância que as
políticas de fomento à industrialização, as inovações no processo produtivo, bem
como a oferta de matérias-primas, e de força de trabalho, têm tido para atrair
unidades industriais para essa região; e, 5) verificamos o aumento de renda
promovido pela indústria, que, aliás, tem contribuído, de certa forma, para uma
melhoria geral das condições de vida da população do Sudoeste do Paraná.
|
117 |
Industrialização e desenvolvimento do sudoeste do Paraná / Industrialization and development of the Paraná Southwest.Flores, Edson Luiz 14 July 2009 (has links)
Made available in DSpace on 2017-05-12T14:42:26Z (GMT). No. of bitstreams: 1
CAPITULO I.pdf: 3435123 bytes, checksum: 3a217406f96215893a8f341510f2421a (MD5)
Previous issue date: 2009-07-14 / How wrote Marx and Engels (1998), the production is very important to man, more
than this, it`s the responsible by own humanization; to measure that the men, to
opposite of the animals that live by nature (offer of foods, resisting by climatic
intemperance etc), get to produce the instrument of survival (foods, clothes, home
etc). However, we verified that in the human sciences, specifically at the Geography,
in the last epochs has been given little attention to the production, mainly to
industrialization. In this study, we realized a analysis about the industry`s importance
at the Southwest of the Parana`s development, region of new population occupation,
happened mainly after the 1940`s. By the realization of that study, proceeded to light
from the historic materialism, method which to consist, thick manner, in to analyses
the formations and transformations occurred at the society from active production
and from work`s connections intrinsic to its. The theoretical reference Marxism
mainly the Karl Marx, Friedrich Engels, Vladimir Lenin, Karl Kautsky, Ignácio Rangel,
Armem Mamigonian`s studies, among others abetted us to that realized a detailed
research, interviewing owns and executives of industrial unities from region, salaried
workers etc, looking at to get information and dies about the thematic studied. In this
discourse, we analyzed: 1) the importance of the industry (mainly the wood`s branch)
to the population of this region from Parana; 2) we verified the conditions that the
social formation (marked by presence of few owns of earth, craftsmen and
merchants) offered to organize a few merchant production; 3) we saw the importance
that had the financial capital, as modernizing the agricultural as financing the
industrial production (mainly from the second half at 1970`s); 4) we saw the
importance that the abetment politics to industrialization, the offer of raw materials,
and of work`s power, have been to attract industrial unities to this region; and, 5) we
verified the increase of income advanced by industry, that besides, has been
contributed, somehow, to a geral advancement of the life`s conditions from
Southwest of the Parana`s population. / Como escreveram Marx e Engels (1998), a produção é de suma importância ao
homem, mais do que isso, ela é responsável pela própria humanização; à medida
que os homens, ao contrário dos animais que vivem à mercê da natureza (da oferta
de alimentos, resistindo às intempéries climáticas etc.), conseguem produzir os
meios de sobrevivência dos quais necessitam (alimentos, vestuário, residências
etc.). No entanto, verificamos que nas ciências humanas, especificamente na
Geografia, nos últimos tempos tem-se dado pouca atenção à produção,
principalmente à industrialização. Nessa pesquisa, realizamos uma análise da
importância da indústria no desenvolvimento do Sudoeste do Paraná, região de
ocupação populacional recente, ocorrida principalmente após a década de 1940.
Para a realização de tal estudo, procedemos à luz do materialismo histórico, método
que consiste, grosso modo, em analisar as formações e transformações ocorridas na
sociedade a partir da organização produtiva e das relações de trabalho inerentes a
ela. O referencial teórico marxista principalmente estudos de Karl Marx, Friedrich
Engels, Vladimir Lenin, Karl Kautsky, Ignácio Rangel, Armem Mamigonian, entre
outros nos apoiou para que realizássemos uma pesquisa minuciosa, entrevistando
proprietários e/ou administradores de unidades industriais da região, trabalhadores
assalariados etc., visando obter informações e dados acerca da temática estudada.
Nessa dissertação, analisamos: 1) a importância da indústria (principalmente o ramo
da madeira) para o povoamento dessa região do Paraná; 2) verificamos as
condições que a formação social (marcada pela presença de pequenos proprietários
de terra, artesãos e comerciantes) ofereceu ao constituir uma pequena produção
mercantil; 3) observamos a importância que o capital financeiro teve, tanto
modernizando a agricultura como financiando a produção industrial (principalmente a
partir da segunda metade da década de 1970); 4) observamos a importância que as
políticas de fomento à industrialização, as inovações no processo produtivo, bem
como a oferta de matérias-primas, e de força de trabalho, têm tido para atrair
unidades industriais para essa região; e, 5) verificamos o aumento de renda
promovido pela indústria, que, aliás, tem contribuído, de certa forma, para uma
melhoria geral das condições de vida da população do Sudoeste do Paraná.
|
118 |
Exact Diagonalization of Few-electron Quantum DotsHakimi, Shirin January 2009 (has links)
<p>We consider a system of few electrons trapped in a two-dimensional circularquantum dot with harmonic confinement and in the presence of ahomogeneous magnetic field, with focus on the role of e-e interaction. Byperforming the exact diagonalization of the Hamiltonian in second quantization,the low-lying energy levels for spin polarized system are obtained. The singlet-triplet oscillation in the ground state of the two-electron system showing up inthe result is explained due to the role of Coulomb interaction. The splitting ofthe lowest Landau level is another effect of the e-e interaction, which is alsoobserved in the results.</p>
|
119 |
Exact Diagonalization of Few-electron Quantum DotsHakimi, Shirin January 2009 (has links)
We consider a system of few electrons trapped in a two-dimensional circularquantum dot with harmonic confinement and in the presence of ahomogeneous magnetic field, with focus on the role of e-e interaction. Byperforming the exact diagonalization of the Hamiltonian in second quantization,the low-lying energy levels for spin polarized system are obtained. The singlet-triplet oscillation in the ground state of the two-electron system showing up inthe result is explained due to the role of Coulomb interaction. The splitting ofthe lowest Landau level is another effect of the e-e interaction, which is alsoobserved in the results.
|
120 |
Contributions à l'apprentissage grande échelle pour la classification d'images / Contributions to large-scale learning for image classificationAkata, Zeynep 06 January 2014 (has links)
La construction d'algorithmes classifiant des images à grande échelle est devenue une t^ache essentielle du fait de la difficulté d'effectuer des recherches dans les immenses collections de données visuelles non-etiquetées présentes sur Internet. L'objetif est de classifier des images en fonction de leur contenu pour simplifier la gestion de telles bases de données. La classification d'images à grande échelle est un problème complexe, de par l'importance de la taille des ensembles de données, tant en nombre d'images qu'en nombre de classes. Certaines de ces classes sont dites "fine-grained" (sémantiquement proches les unes des autres) et peuvent même ne contenir aucun représentant étiqueté. Dans cette thèse, nous utilisons des représentations à l'état de l'art d'images et nous concentrons sur des méthodes d'apprentissage efficaces. Nos contributions sont (1) un banc d'essai d'algorithmes d'apprentissage pour la classification à grande échelle et (2) un nouvel algorithme basé sur l'incorporation d'étiquettes pour apprendre sur des données peu abondantes. En premier lieu, nous introduisons un banc d'essai d'algorithmes d'apprentissage pour la classification à grande échelle, dans un cadre entièrement supervisé. Il compare plusieurs fonctions objectifs pour apprendre des classifieurs linéaires, tels que "un contre tous", "multiclasse", "classement", "classement avec pondération" par descente de gradient stochastique. Ce banc d'essai se conclut en un ensemble de recommandations pour la classification à grande échelle. Avec une simple repondération des données, la stratégie "un contre tous" donne des performances meilleures que toutes les autres. Par ailleurs, en apprentissage en ligne, un pas d'apprentissage assez petit s'avère suffisant pour obtenir des résultats au niveau de l'état de l'art. Enfin, l'arrêt prématuré de la descente de gradient stochastique introduit une régularisation qui améliore la vitesse d'entraînement ainsi que la capacité de régularisation. Deuxièmement, face à des milliers de classes, il est parfois difficile de rassembler suffisamment de données d'entraînement pour chacune des classes. En particulier, certaines classes peuvent être entièrement dénuées d'exemples. En conséquence, nous proposons un nouvel algorithme adapté à ce scénario d'apprentissage dit "zero-shot". Notre algorithme utilise des données parallèles, comme les attributs, pour incorporer les classes dans un espace euclidien. Nous introduisons par ailleurs une fonction pour mesurer la compatibilité entre image et étiquette. Les paramètres de cette fonction sont appris en utilisant un objectif de type "ranking". Notre algorithme dépasse l'état de l'art pour l'apprentissage "zero-shot", et fait preuve d'une grande flexibilité en permettant d'incorporer d'autres sources d'information parallèle, comme des hiérarchies. Il permet en outre une transition sans heurt du cas "zero-shot" au cas où peu d'exemples sont disponibles. / Building algorithms that classify images on a large scale is an essential task due to the difficulty in searching massive amount of unlabeled visual data available on the Internet. We aim at classifying images based on their content to simplify the manageability of such large-scale collections. Large-scale image classification is a difficult problem as datasets are large with respect to both the number of images and the number of classes. Some of these classes are fine grained and they may not contain any labeled representatives. In this thesis, we use state-of-the-art image representations and focus on efficient learning methods. Our contributions are (1) a benchmark of learning algorithms for large scale image classification, and (2) a novel learning algorithm based on label embedding for learning with scarce training data. Firstly, we propose a benchmark of learning algorithms for large scale image classification in the fully supervised setting. It compares several objective functions for learning linear classifiers such as one-vs-rest, multiclass, ranking and weighted average ranking using the stochastic gradient descent optimization. The output of this benchmark is a set of recommendations for large-scale learning. We experimentally show that, online learning is well suited for large-scale image classification. With simple data rebalancing, One-vs-Rest performs better than all other methods. Moreover, in online learning, using a small enough step size with respect to the learning rate is sufficient for state-of-the-art performance. Finally, regularization through early stopping results in fast training and a good generalization performance. Secondly, when dealing with thousands of classes, it is difficult to collect sufficient labeled training data for each class. For some classes we might not even have a single training example. We propose a novel algorithm for this zero-shot learning scenario. Our algorithm uses side information, such as attributes to embed classes in a Euclidean space. We also introduce a function to measure the compatibility between an image and a label. The parameters of this function are learned using a ranking objective. Our algorithm outperforms the state-of-the-art for zero-shot learning. It is flexible and can accommodate other sources of side information such as hierarchies. It also allows for a smooth transition from zero-shot to few-shots learning.
|
Page generated in 0.0774 seconds