Log In Sign Up

Attend, Infer, Repeat: Fast Scene Understanding with Generative Models

by   S. M. Ali Eslami, et al.

We present a framework for efficient inference in structured image models that explicitly reason about objects. We achieve this by performing probabilistic inference using a recurrent neural network that attends to scene elements and processes them one at a time. Crucially, the model itself learns to choose the appropriate number of inference steps. We use this scheme to learn to perform inference in partially specified 2D models (variable-sized variational auto-encoders) and fully specified 3D models (probabilistic renderers). We show that such models learn to identify multiple objects - counting, locating and classifying the elements of a scene - without any supervision, e.g., decomposing 3D images with various numbers of objects in a single forward pass of a neural network. We further show that the networks produce accurate inferences when compared to supervised counterparts, and that their structure leads to improved generalization.


page 5

page 6

page 7

page 12

page 13

page 14

page 16


Auto-Encoding Sequential Monte Carlo

We introduce AESMC: a method for using deep neural networks for simultan...

Scalable Bayesian Inference for Detection and Deblending in Astronomical Images

We present a new probabilistic method for detecting, deblending, and cat...

Knowledge-Guided Object Discovery with Acquired Deep Impressions

We present a framework called Acquired Deep Impressions (ADI) which cont...

ProbNeRF: Uncertainty-Aware Inference of 3D Shapes from 2D Images

The problem of inferring object shape from a single 2D image is undercon...

Towards a Neural Statistician

An efficient learner is one who reuses what they already know to tackle ...

Auto-Encoder Neural Network Incorporating X-Ray Fluorescence Fundamental Parameters with Machine Learning

We consider energy-dispersive X-ray Fluorescence (EDXRF) applications wh...

Code Repositories


A Tensorfflow implementation of Attend, Infer, Repeat

view repo