Deep Convolutional Inverse Graphics Network

Kulkarni, Tejas; Whitney, W. Bevan; Kohli, Pushmeet; Tenenbaum, Joshua B.

doi:10.48550/arxiv.1503.03167

Cited by 16 publications

(20 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Researchers have encouraged invariance or forgetting with VAEs by utilizing various combinations of weak supervision [1], [23], adversarial training [6], and by trading off reconstruction error against encoding capacity according to the information bottleneck principle [25]. Others have sought to fully disentangle the latent space into independent partitions [14], thereby enabling the isolation of wanted from unwanted information. Whilst unsupervised methods for disentanglement exist [4], [5], recent research indicates that such methods do not achieve disentanglement consistently [19] and, even if they do, they do not allocate factors to the latent space in a predictable way.…”

Section: Literature Reviewmentioning

confidence: 99%

Improving Robot Localisation by Ignoring Visual Distraction

Méndez¹,

Vowels²,

Bowden³

2021

Preprint

View full text Add to dashboard Cite

Attention is an important component of modern deep learning. However, less emphasis has been put on its inverse: ignoring distraction. Our daily lives require us to explicitly avoid giving attention to salient visual features that confound the task we are trying to accomplish. This visual prioritisation allows us to concentrate on important tasks while ignoring visual distractors.In this work, we introduce Neural Blindness, which gives an agent the ability to completely ignore objects or classes that are deemed distractors. More explicitly, we aim to render a neural network completely incapable of representing specific chosen classes in its latent space. In a very real sense, this makes the network "blind" to certain classes, allowing and agent to focus on what is important for a given task, and demonstrates how this can be used to improve localisation.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Improving Robot Localisation by Ignoring Visual Distraction

Méndez¹,

Vowels²,

Bowden³

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Kulkarni et al [15] present an inverse graphics model that disentangles pose and lighting. They propose a novel training procedure for factoring their latent space.…”

Section: Related Workmentioning

confidence: 99%

From A to Z: Supervised Transfer of Style and Content Using Deep Neural Network Generators

Upchurch,

Snavely,

Bala

2016

Preprint

View full text Add to dashboard Cite

We propose a new neural network architecture for solving single-image analogies-the generation of an entire set of stylistically similar images from just a single input image. Solving this problem requires separating image style from content. Our network is a modified variational autoencoder (VAE) that supports supervised training of singleimage analogies and in-network evaluation of outputs with a structured similarity objective that captures pixel covariances. On the challenging task of generating a 62-letter font from a single example letter we produce images with 22.4% lower dissimilarity to the ground truth than state-of-the-art.

show abstract

“…Many equivalent parametrisations of the manifold may be possible, some allowing a more semantic interpretation of each latent coefficient than others, and the standard AE or DAE does not distinguish among these parametrisations. There has been recent interest in adapting the training schemes of autoencoders such that the latent representation is explicitly semantic, capturing attributes of the input in specific subsets of the latent variables [3,4]. We refer to these as partitioned autoencoders since the latent variables are partitioned into subsets which are treated differently from each other during training.…”

Section: Partitioned Autoencodersmentioning

confidence: 99%

“…We refer to these as partitioned autoencoders since the latent variables are partitioned into subsets which are treated differently from each other during training. Crucially, in this prior work the training scheme relies heavily on the existence of large structured datasets: in [4] a balanced dataset of labelled digit images; in [3] a dataset of faces constructed through systematic variation of attributes such as pose and lighting. Without such known structure in the training data, their proposed training schemes will be either impossible to apply, or biased by the presence of unbalanced or correlated factors in the training data.…”

Section: Partitioned Autoencodersmentioning

confidence: 99%

Denoising without access to clean data using a partitioned autoencoder

Stowell,

Turner

2015

Preprint

View full text Add to dashboard Cite

Training a denoising autoencoder neural network requires access to truly clean data, a requirement which is often impractical. To remedy this, we introduce a method to train an autoencoder using only noisy data, having examples with and without the signal class of interest. The autoencoder learns a partitioned representation of signal and noise, learning to reconstruct each separately. We illustrate the method by denoising birdsong audio (available abundantly in uncontrolled noisy datasets) using a convolutional autoencoder.

show abstract

Deep Convolutional Inverse Graphics Network

Cited by 16 publications

References 0 publications

Improving Robot Localisation by Ignoring Visual Distraction

Improving Robot Localisation by Ignoring Visual Distraction

From A to Z: Supervised Transfer of Style and Content Using Deep Neural Network Generators

Denoising without access to clean data using a partitioned autoencoder

Contact Info

Product

Resources

About