Sébastien Ehrhardt
YOU?
Author Swipe
View article: LSD-C: Linearly Separable Deep Clusters
LSD-C: Linearly Separable Deep Clusters Open
We present LSD-C, a novel method to identify clusters in an unlabeled dataset. Our algorithm first establishes pairwise connections in the feature space between the samples of the minibatch based on a similarity metric. Then it regroups in…
View article: AutoNovel: Automatically Discovering and Learning Novel Visual Categories
AutoNovel: Automatically Discovering and Learning Novel Visual Categories Open
We tackle the problem of discovering novel classes in an image collection given labelled examples of other classes. We present a new approach called AutoNovel to address this problem by combining three ideas: (1) we suggest that the common…
View article: Co-Attention for Conditioned Image Matching
Co-Attention for Conditioned Image Matching Open
We propose a new approach to determine correspondences between image pairs in the wild under large changes in illumination, viewpoint, context, and material. While other approaches find correspondences between pairs of images by treating t…
View article: 3D Multi-bodies: Fitting Sets of Plausible 3D Human Models to Ambiguous Image Data
3D Multi-bodies: Fitting Sets of Plausible 3D Human Models to Ambiguous Image Data Open
We consider the problem of obtaining dense 3D reconstructions of humans from single and partially occluded views. In such cases, the visual evidence is usually insufficient to identify a 3D reconstruction uniquely, so we aim at recovering …
View article: D2D: Learning to find good correspondences for image matching and manipulation
D2D: Learning to find good correspondences for image matching and manipulation Open
We propose a new approach to determining correspondences between image pairs under large changes in illumination, viewpoint, context, and material. While most approaches seek to extract a set of reliably detectable regions in each image wh…
View article: RELATE: Physically Plausible Multi-Object Scene Synthesis Using Structured Latent Spaces
RELATE: Physically Plausible Multi-Object Scene Synthesis Using Structured Latent Spaces Open
We present RELATE, a model that learns to generate physically plausible scenes and videos of multiple interacting objects. Similar to other generative approaches, RELATE is trained end-to-end on raw, unlabeled data. RELATE combines an obje…
View article: Semi-Supervised Learning with Scarce Annotations
Semi-Supervised Learning with Scarce Annotations Open
While semi-supervised learning (SSL) algorithms provide an efficient way to make use of both labelled and unlabelled data, they generally struggle when the number of annotated samples is very small. In this work, we consider the problem of…
View article: Automatically Discovering and Learning New Visual Categories with Ranking Statistics
Automatically Discovering and Learning New Visual Categories with Ranking Statistics Open
We tackle the problem of discovering novel classes in an image collection given labelled examples of other classes. This setting is similar to semi-supervised learning, but significantly harder because there are no labelled examples for th…
View article: RELATE: Physically Plausible Multi-Object Scene Synthesis Using Structured Latent Spaces
RELATE: Physically Plausible Multi-Object Scene Synthesis Using Structured Latent Spaces Open
We present RELATE, a model that learns to generate physically plausible scenes and videos of multiple interacting objects. Similar to other generative approaches, RELATE is trained end-to-end on raw, unlabeled data. RELATE combines an obje…
View article: Small Steps and Giant Leaps: Minimal Newton Solvers for Deep Learning
Small Steps and Giant Leaps: Minimal Newton Solvers for Deep Learning Open
We propose a fast second-order method that can be used as a drop-in replacement for current deep learning solvers. Compared to stochastic gradient descent (SGD), it only requires two additional forward-mode automatic differentiation operat…
View article: Unsupervised Intuitive Physics from Past Experiences
Unsupervised Intuitive Physics from Past Experiences Open
We are interested in learning models of intuitive physics similar to the ones that animals use for navigation, manipulation and planning. In addition to learning general physical principles, however, we are also interested in learning ``on…
View article: Deep Industrial Espionage
Deep Industrial Espionage Open
The theory of deep learning is now considered largely solved, and is well understood by researchers and influencers alike. To maintain our relevance, we therefore seek to apply our skills to under-explored, lucrative applications of this t…
View article: Taking visual motion prediction to new heightfields
Taking visual motion prediction to new heightfields Open
While the basic laws of Newtonian mechanics are well understood, explaining a physical scenario still requires manually modeling the problem with suitable equations and estimating the associated parameters. In order to be able to leverage …
View article: Learning to Represent Mechanics via Long-term Extrapolation and Interpolation
Learning to Represent Mechanics via Long-term Extrapolation and Interpolation Open
While the basic laws of Newtonian mechanics are well understood, explaining a physical scenario still requires manually modeling the problem with suitable equations and associated parameters. In order to adopt such models for artificial in…
View article: Stopping GAN Violence: Generative Unadversarial Networks
Stopping GAN Violence: Generative Unadversarial Networks Open
While the costs of human violence have attracted a great deal of attention from the research community, the effects of the network-on-network (NoN) violence popularised by Generative Adversarial Networks have yet to be addressed. In this w…
View article: Learning A Physical Long-term Predictor
Learning A Physical Long-term Predictor Open
Evolution has resulted in highly developed abilities in many natural intelligences to quickly and accurately predict mechanical phenomena. Humans have successfully developed laws of physics to abstract and model such mechanical phenomena. …