OpenDR: An Approximate Differentiable Renderer

By Matthew Maverick Loper, Michael J. Black. 2014

572 citations recorded by the source corpus.

Subfields: Computational Mechanics, Computer Graphics and Computer-Aided Design, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Mechanics

In Computer Graphics and Computer-Aided Design

In Computer Vision and Pattern Recognition

Citations in the corpus

Learning To Simulate (2018, background): “However, for applications like computer vision, differentiable rendering (Loper & Black, 2014) cannot yet compete with more mature rendering engines in terms of data ﬁdelity and expressive power.”
RenderGAN: Generating Realistic Labeled Data (2016, background): “While there exists differentiable renders (Loper and Black, 2014), we found it the easiest to train a neural network to emulate the 3D model.”
Generalization in anti-causal learning (2018, methodology): “In computer vision, 3D render engines have successfully been used in a line of research known as vision as inverse graphics to invert the physical process that produced an image (Loper and Black, 2014; Eslami et al., 2016).”
3D Shape Induction from 2D Views of Multiple Objects (2016, positive): “…are commonly used with rendering pipelines ( e.g ., OpenGL [31]) and can also be adjusted to match image statistics using a differentiable ren-derer ( e.g ., OpenDR [20]), but the variability of the mesh topology makes it difﬁcult to generate them in a consistent manner across instances.”
Picture: A probabilistic programming language for scene perception (2015, methodology): “Inspired by the differentiable renderer [29], Picture also supports expressing AR’s entire graphics pipeline as Picture code, enabling the language to express end-to-end differentiable generative models.”
Unsupervised Learning of Depth and Ego-Motion from Video (2017, background): “Several authors have explored building differentiable rendering operations into their models that are trained in this way, e.g., [19, 29, 34].”
Neural 3D Mesh Renderer (2017, neutral): “OpenDR [15] is a differentiable renderer.”
Neural Scene De-rendering (2017, background): “This is technically nontrivial because graphics engines are often not differentiable, with few exceptions [22].”
Synthesizing Programs for Images using Reinforced Adversarial Learning (2018, background): “The idea of inverting simulators to interpret images has been explored extensively in recent years (Nair et al., 2008; Paysan et al., 2009; Mansinghka et al., 2013; Loper & Black, 2014; Kulkarni et al., 2015a; Jampani et al., 2015).”
Adversarial Geometry and Lighting using a Differentiable Renderer (2018, background): “Figure 7: Our approach faster and more scalable comparing to previous differentiable renderer [24].”
AutoFlow: Learning a Better Training Set for Optical Flow (2021, background): “Differentiable rendering [29, 37] enables gradients to be passed to rendering parameters, which, however, do not directly relate to the scene distribution hyperparameters.”
Deep Generative Vision as Approximate Bayesian Computation (2014, background): “Probabilistic formulations of inverse graphics have recently been proposed for a variety of 2D and 3D vision problems [15, 12, 14, 9].”
Pix2Shape: Towards Unsupervised Learning of 3D Scenes from Images Using a View-Based Representation (2020, methodology): “OpenDR (Loper and Black 2014), as used by Henderson and Ferrari (2018), applies first order Taylor approximation to compute gradients.”
Small Data Challenges in Big Data Era: A Survey of Recent Progress on Unsupervised and Semi-Supervised Methods (2019, unknown): “…construct the 3D representation, and transforming auto-encoders [60], [89] that use a domain-speciﬁc decoder to reconstruct images, as well as [90] with an approximate differentiable renderer to explicitly capture the relationship between changes in model parameters and image observations.”
Soft Rasterizer: Differentiable Rendering for Unsupervised Single-View Mesh Reconstruction (2019, background): “More recently, Loper and Black [26] further introduce OpenDR, an approximate differentiable renderer which can be incorporated into probabilistic programming framework to obtain derivatives with respect to the model parameters.”