A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2019; you can also visit the original URL.
The file type is application/pdf.
Filters
Transfer of View-manifold Learning to Similarity Perception of Novel Objects
[article]
2017
arXiv
pre-print
Preserved Fulltext
We develop a model of perceptual similarity judgment based on re-training a deep convolution neural network (DCNN) that learns to associate different views of each 3D object to capture the notion of object ...
We found that this ability is not limited to the trained objects, but transfers to novel objects in both trained and untrained categories, as well as to a variety of completely novel artificial synthetic ...
Such a clean setup is necessary for us to study the effect of object persistency constraint on novel objects, as well as the transferability of view-manifold learning to similarity perception. ...
arXiv:1704.00033v1
fatcat:km5kdd34lvhqtfsvxt4pceg2xy
Model Guided Multimodal Imaging and Visualization for Computer Assisted Interventions
2011
IAPR International Workshop on Machine Vision Applications
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
In this short paper, we 1 try to present a summary of some of our activities relevant to computer assisted interventions and computer vision. ...
It was not possible to cover all aspects of our research within this paper, but we hope to provide an overview on some of these within this short paper. ...
In [54] , we apply manifold learning to the multi-modal registration problem. ...
dblp:conf/mva/Navab11
fatcat:mwayyz5jlfhefmlx2s2d7if5yi
Toward the neural implementation of structure learning
2016
Current Opinion in Neurobiology
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2016; you can also visit the original URL.
The file type is application/pdf.
Animals categorize objects, learn to vocalize and may even estimate causal relationships -all in the face of data that is often ambiguous and sparse. ...
Towards the neural implementation of structure learning Tervo, Tenenbaum and Gershman 105 www.sciencedirect.com ...
In principle, transfer learning can be achieved without induction of an abstract rule: a simpler solution, such as the learning of a good way to group perceptions, can facilitate performance in a similar ...
doi:10.1016/j.conb.2016.01.014
pmid:26874471
fatcat:z5jf2k4tc5aozkajjwgqe2twfm
The Structure Transfer Machine Theory and Applications
[article]
2019
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Note that this fulltext copy is not of the "primary" version of this work. The version it corresponds to is:
2018 pre-print arXiv:1804.00243v1 fatcat:v2sjanjgfncwrbjiiabyc6piti
We theoretically show that such an expected value of the representation (mean) is achievable if the manifold structure can be transferred from the data space to the feature space. ...
We propose a new representation learning method, termed Structure Transfer Machine (STM), which enables feature learning process to converge at the representation expectation in a probabilistic way. ...
Upon such a proven, we present a novel structure transfer machine (STM) to learn structured deep features, the framework of which is illustrated in Fig. 1 . ...
arXiv:1804.00243v2
fatcat:dmlnobq75rf3jomd5xxrr7kkgu
Generative Visual Manipulation on the Natural Image Manifold
[article]
2018
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2019; you can also visit the original URL.
The file type is application/pdf.
Other Versions
We then define a class of image editing operations, and constrain their output to lie on that learned manifold at all times. ...
In this paper, we propose to learn the natural image manifold directly from data using a generative adversarial neural network. ...
However, we expect a model trained on many categories (e.g. 1, 000) to generalize better to novel objects. ...
arXiv:1609.03552v3
fatcat:4p4nhm2uhjbpdfzekiijr7utfa
Distal Attribution and Distance Perception in Sensory Substitution
2010
Perception
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2017; you can also visit the original URL.
The file type is application/pdf.
Performance transferred to the non-dominant arm and to a rotated body orientation, demonstrating that learning did not depend on a joint-specific sensorimotor relationship between target distance and arm ...
To investigate how it arises would require manipulating the usual sensory interface in order to present novel proximal stimulation and analyze the emergence of distal awareness. ...
We would like to thank Elias Jaffa, Rajesh Shah, Max DiLuca for their assistance, Charles Lenay, Roger Cholewiak, Jack Loomis, and Kevin O'Regan for their advice. ...
doi:10.1068/p6366
pmid:20402243
pmcid:PMC3780418
fatcat:v4dzfulkxbdh3kjojvvadlzh5a
Transferring Colours to Grayscale Images by Locally Linear Embedding
2008
Procedings of the British Machine Vision Conference 2008
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2009; you can also visit the original URL.
The file type is application/pdf.
Then we synthesize the objective chromatic channels using the learned relations. Experiments show that our method gives superior results to those of the previous work. ...
In contrast to many previous computer-aided colourizing methods, which require intensive and accurate human intervention, our method needs only the user to provide a colourful image of the similar content ...
similar manifolds. ...
doi:10.5244/c.22.83
dblp:conf/bmvc/LiHZ08
fatcat:bmqv43hkmzc4zc7ggrz5f5xcpy
Deep Predictive Policy Training using Reinforcement Learning
[article]
2017
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2019; you can also visit the original URL.
The file type is application/pdf.
The architecture consists of three sub-networks referred to as the perception, policy and behavior super-layers. ...
Skilled robot task learning is best implemented by predictive action policies due to the inherent latency of sensorimotor processes. ...
We will also study transferability of features extracted at different layers of the model to extend the network to manipulate multiple target objects without necessarily re-training from scratch. ...
arXiv:1703.00727v1
fatcat:iuy4lllozvb5zm6hchydrshxmm
On the Intrinsic Limits to Representationally-Adaptive Machine-Learning
[article]
2015
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
We shall argue in this paper, however, that a full realization of this notion requires that, in addition to the capacity to adapt to novel data, autonomous online learning must ultimately incorporate the ...
Online learning is a familiar problem setting within Machine-Learning in which data is presented serially in time to a learning agent, requiring it to progressively adapt within the constraints of the ...
to map increasing numbers of the original percepts on to a singular novel percept). ...
arXiv:1503.02626v1
fatcat:7v4i3ml5uzfuzit34c7wlvyo2u
MSNet: A Deep Multi-scale Submanifold Network for Visual Classification
[article]
2022
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Other Versions
Although there are many different attempts to develop effective deep architectures for data processing on the Riemannian manifold of SPD matrices, a very few solutions explicitly mine the local geometrical ...
Accordingly, in this work we propose an SPD network designed with this objective in mind. ...
53] , Novel View [54] , Transition Forests (TF) [55] , Temporal Convolutional Network (TCN) [56] , LSTM [47] and Unified Hand and Object Model [57] . ...
arXiv:2201.10145v2
fatcat:pvgkpoatbrburdeqdzqb4nfoty
Guest Editors' Introduction to the Special Section on Learning with Shared Information for Computer Vision and Multimedia Analysis
2018
IEEE Transactions on Pattern Analysis and Machine Intelligence
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2019; you can also visit the original URL.
The file type is application/pdf.
Shao, introduces the concept of hetero-manifold for integrating the uniand cross-modal similarities of multi-modal data in a global view. ...
Corso, proposes a novel model representing bimodal percepts that exploits the compositional structure of language. ...
doi:10.1109/tpami.2018.2804998
fatcat:urin3tvgy5f7ng5djfvlm4mop4
Unsupervised Feature Learning of Human Actions as Trajectories in Pose Embedding Manifold
[article]
2018
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
We demonstrate state-of-the-art transfer-ability of the learned representation against other supervisedly and unsupervisedly learned motion embeddings for the task of fine-grained action recognition on ...
In order to realize a continuous pose embedding manifold with improved reconstructions, we propose an unsupervised, manifold learning procedure named Encoder GAN, (or EnGAN). ...
Approach In this section we describe the proposed pose manifold learning methodology along with the carefully carried out preprocessing steps to make EnGAN invariant to translation, view and scale, both ...
arXiv:1812.02592v1
fatcat:37jnz4444faudnvaiao5d6sjym
Generative Visual Manipulation on the Natural Image Manifold
[chapter]
2016
Lecture Notes in Computer Science
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2017; you can also visit the original URL.
The file type is application/pdf.
We then define a class of image editing operations, and constrain their output to lie on that learned manifold at all times. ...
In this paper, we propose to learn the natural image manifold directly from data using a generative adversarial neural network. ...
However, we expect a model trained on many categories (e.g. 1, 000) to generalize better to novel objects. ...
doi:10.1007/978-3-319-46454-1_36
fatcat:y64u4pnkxng3jnlogi7j2p443y
A Transfer Learning Approach to Cross-Modal Object Recognition: From Visual Observation to Robotic Haptic Exploration
2019
IEEE Transactions on robotics
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Note that this fulltext copy is not of the "primary" version of this work. The version it corresponds to is:
2020 pre-print arXiv:2001.06673v1 fatcat:uthlse4alja5dbzg4oc2tffnvi
With this term, we mean that the robot observes a set of objects with visual perception and, later on, it is able to recognize such objects only with tactile exploration, without having touched any object ...
To tackle this problem, we propose an approach constituted by four steps: finding a visuo-tactile common representation, defining a suitable set of features, transferring the features across the domains ...
With respect to the work in [3] , the novel additional research questions are • do transfer learning approaches help to improve the performance of cross-modal visuo-tactile problems? ...
doi:10.1109/tro.2019.2914772
fatcat:jlcmek2rsrb77ndiorpn7ks6ri
Discovering Latent States for Model Learning: Applying Sensorimotor Contingencies Theory and Predictive Processing to Model Context
[article]
2016
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
How can we enable artificial agents, such as robots, to acquire some form of generic knowledge, which they could leverage for the learning of new skills? ...
This paper argues that, like the brain, the cognitive system of artificial agents has to develop a world model to support adaptive behavior and learning. ...
Taking again the example of object perception, we could say that a given e k corresponds to an agent-environment configuration where a certain object is in front of the agent (as compared to the object ...
arXiv:1608.00359v1
fatcat:xasdljbsafdqjpsunv6shqbwhu
« Previous
Showing results 1 — 15 out of 9,174 results