A Parallel Computation that Assigns Canonical Object-Based Frames of Reference

By Geoffrey E. Hinton. 1981

How researchers see this paper

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

In Neuroscience

Citations in the corpus

Spatial Transformer Networks (2015, background): “In this section we discuss the prior work related to the paper, covering the central ideas of modelling transformations with neural networks [12, 13, 27], learning and analysing transformation-invariant representations [3, 5, 8, 17, 19, 25], as well as attention and detection mechanisms for feature ”
Dynamic Routing Between Capsules (2017, background): “Olshausen et al. [1993] improves upon Hinton [1981b] dynamic connections and presents a biologically plausible, position and scale invariant model of object representations.”
How to Represent Part-Whole Hierarchies in a Neural Network (2021, background): “If we allow three-way interactions in which the activity of one neuron gates the connection between two other neurons [Hinton, 1981c], it is easy to make the connections dynamic, but it is still unclear how to dynamically create a graph structure without the ability to allocate neurons on the ﬂy.”
A neurobiological model of visual attention and invariant pattern recognition based on dynamic routing of information (1993, background): “In machine vision, this is known as the viewpoint consistency constraint, and it has proved to be a powerful computational strategy for object recognition systems (Hinton, 1981b; Lowe, 1987).”
Learning to Relate Images (2011, background): “We discuss how the common computational principle behind all these methods are multiplicative interactions , which were introduced to the vision community 30 years ago under the terms “mapping units” [18] and “dynamic mappings” [48].”
Dynamic binding in a neural network for shape recognition. (1992, methodology): “Feature matching models extract diagnostic features from an image and use those as the basis for recognition (e.g., Hinton, 1981; Hummel, Biederman, Gerhardstein, & Hilton, 1988; Lindsay & Norman, 1977; Selfridge, 1959; Selfridge & Neisser, 1960).”
Deforming Autoencoders: Unsupervised Disentangling of Shape and Appearance (2018, background): “A thread of works has been aimed at dynamically rerouting the processing of information within the network’s graph based on the input, starting from neural computation arguments [21,22,23] and eventually translating into concrete algorithms, such as the ‘capsule’ works of [24,25] that bind neurons o”
Batch-shaping for learning conditional channel gated networks (2019, methodology): “ as Convnet-AIG [38], SkipNet [39] and soft-guided adaptively-dropped neural network [40]. 2 Background and related work Literature on gating connections for deep neural networks dates back to Hinton [12]. Gating can be seen as a tri-way connection in a neural network [7], where one output can only ”
Deep learning for AI (2021, background): “While multiplicative gating is an old idea for such things as coordinate transforms 44 and powerful forms of recurrent networks, 52 its recent forms have made it mainstream.”
Deep learning for cardiovascular medicine: a practical primer. (2019, background): “Capsule-based neural network, 32,33 meta-learning (learning how the network learns), 34–36 DeepMath (learn mathematical proofs), 37 or self-play (two agents learn by win and loss) 34 S2 ).”
Tensor Product Variable Binding and the Representation of Symbolic Structures in Connectionist Systems (1990, background): “product occurs not at the unit but at the junction of two connections; the two activities entering the triangular junction [12] from the filler and role units are multiplied together and the result is sent along the third line to the binding unit.”
Learning to Segment Images Using Dynamic Feature Binding (1991, methodology): “Further, the segmentation-recognition cycle can iterate, allowing the recognition system to propose refinements of the initial segmentation, which in turn refines the output of the recognition system (Hinton 1981; Hanson and Riseman 1978; Waltz 1975).”