Deep Learning for Vanishing Point Detection Using an Inverse Gnomonic Projection

Kluger, Florian; Ackermann, Hanno; Yang, Michael Ying; Rosenhahn, Bodo

doi:10.1007/978-3-319-66709-6_2

Cited by 40 publications

(44 citation statements)

References 30 publications

(69 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We are also interested in making our method non-deterministic, similar to a Markov Chain Neural Network [1] and we want to adapt our method to other data sets in 3D [33]. We further expect that problems, such as vanishing point estimation [21] or semantic image understanding [35] can benefit from our approach.…”

Section: Discussionmentioning

confidence: 99%

Learning Disentangled Representations via Independent Subspaces

Awiszus

Ackermann

Rosenhahn

2019

2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)

Self Cite

View full text Add to dashboard Cite

Image generating neural networks are mostly viewed as black boxes, where any change in the input can have a number of globally effective changes on the output. In this work, we propose a method for learning disentangled representations to allow for localized image manipulations. We use face images as our example of choice. Depending on the image region, identity and other facial attributes can be modified. The proposed network can transfer parts of a face such as shape and color of eyes, hair, mouth, etc. directly between persons while all other parts of the face remain unchanged. The network allows to generate modified images which appear like realistic images. Our model learns disentangled representations by weak supervision. We propose a localized resnet autoencoder optimized using several loss functions including a loss based on the semantic segmentation, which we interpret as masks, and a loss which enforces disentanglement by decomposition of the latent space into statistically independent subspaces. We evaluate the proposed solution w.r.t. disentanglement and generated image quality. Convincing results are demonstrated using the CelebA dataset [24].

show abstract

Section: Discussionmentioning

confidence: 99%

Learning Disentangled Representations via Independent Subspaces

Awiszus

Ackermann

Rosenhahn

2019

2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)

Self Cite

View full text Add to dashboard Cite

show abstract

“…The horizon can serve as a cue in image understanding [52] or for image editing [25]. Traditionally, this task is solved via vanishing point detection and geometric reasoning [37,24,57,42], often assuming a Manhattan or Atlanta world. We take a simpler approach and use a general purpose CNN that predicts a set of 64 2D points based on the image to which we fit a line with RANSAC, see Fig.…”

Section: Horizon Linesmentioning

confidence: 99%

Neural-Guided RANSAC: Learning Where to Sample Model Hypotheses

Brachmann

Rother

2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

226

171

View full text Add to dashboard Cite

We present Neural-Guided RANSAC (NG-RANSAC), an extension to the classic RANSAC algorithm from robust optimization. NG-RANSAC uses prior information to improve model hypothesis search, increasing the chance of finding outlier-free minimal sets. Previous works use heuristic side information like hand-crafted descriptor distance to guide hypothesis search. In contrast, we learn hypothesis search in a principled fashion that lets us optimize an arbitrary task loss during training, leading to large improvements on classic computer vision tasks. We present two further extensions to NG-RANSAC. Firstly, using the inlier count itself as training signal allows us to train neural guidance in a self-supervised fashion. Secondly, we combine neural guidance with differentiable RANSAC to build neural networks which focus on certain parts of the input data and make the output predictions as good as possible. We evaluate NG-RANSAC on a wide array of computer vision tasks, namely estimation of epipolar geometry, horizon line estimation and camera re-localization. We achieve superior or competitive results compared to state-of-the-art robust estimators, including very recent, learned ones.

show abstract

“…In this paper we do not discuss the different variants of neural networks and their possibilities for optimization, autoencoders [17], incremental learning [14,15] or data management [4]. We only want to state, that neural networks are commonly used for competing in different benchmarks with remarkable performance [13,21,11]. For this work it is only important to clarify that a neural network is a deterministic function and in its nature not suited for modeling Markov chains.…”

Section: Neural Networkmentioning

confidence: 99%

Markov Chain Neural Networks

Awiszus

Rosenhahn

2018

2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

Self Cite

View full text Add to dashboard Cite

In this work we present a modified neural network model which is capable to simulate Markov Chains. We show how to express and train such a network, how to ensure given statistical properties reflected in the training data and we demonstrate several applications where the network produces non-deterministic outcomes. One example is a random walker model, e.g. useful for simulation of Brownian motions or a natural Tic-Tac-Toe network which ensures non-deterministic game behavior. Foundations and Stochastic Neural NetworksThis section summarizes the foundations and some stateof-the art for this work with a focus on Markov Chains and Neural Networks. Graphical ModelsA probabilistic graphical model (PGM) is a probabilistic model for which a graph G = (V, E) with given edges

show abstract

Deep Learning for Vanishing Point Detection Using an Inverse Gnomonic Projection

Cited by 40 publications

References 30 publications

Learning Disentangled Representations via Independent Subspaces

Learning Disentangled Representations via Independent Subspaces

Neural-Guided RANSAC: Learning Where to Sample Model Hypotheses

Markov Chain Neural Networks

Contact Info

Product

Resources

About