I first heard about GPGP in this talk by Josh Tenenbaum: Two architectures for one-shot learning
It is similar to the idea of analysis by synthesis / Helmholtz machines:
GPGP is what happens when you take those ideas and throw probabilistic programming into the mix (as Josh Tenenbaum tends to do). For inference, they use the Metropolis-Hastings algorithm.
It may be a good idea to generalize and throw away the “graphics” part, to have generic “Generative Probabilistic Programs”. But let’s not get ahead of ourselves. Limiting the study to graphics programs seems a good way to first see is this idea works at all.
Quoting the abstract from the GPGP paper:
The idea of computer vision as the Bayesian inverse problem to computer graphics has a long history and an appealing elegance, but it has proved difficult to directly implement. Instead, most vision tasks are approached via complex bottom-up processing pipelines. Here we show that it is possible to write short, simple probabilistic graphics programs that define flexible generative models and to automatically invert them to interpret real-world images. Generative probabilistic graphics programs consist of a stochastic scene generator, a renderer based on graphics software, a stochastic likelihood model linking the renderer’s output and the data, and latent variables that adjust the fidelity of the renderer and the tolerance of the likelihood model. Representations and algorithms from computer graphics, originally designed to produce high-quality images, are instead used as the deterministic backbone for highly approximate and stochastic generative models. This formulation combines probabilistic programming, computer graphics, and approximate Bayesian computation, and depends only on general-purpose, automatic inference techniques. We describe two applications: reading sequences of degraded and adversarially obscured alphanumeric characters, and inferring 3D road models from vehicle-mounted camera images. Each of the probabilistic graphics programs we present relies on under 20 lines of probabilistic code, and supports accurate, approximately Bayesian inferences about ambiguous real-world images.