Principles of Object Perception

By Elizabeth S. Spelke. 1990

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

What can the brain teach us about building artificial intelligence? (2016, background): “iples of persistence, continuity, cohesion and solidity. Young infants believe objects should move along smooth paths, not wink in and out of existence, not inter-penetrate and not act at a distance (Spelke, 1990; Spelke, Gutheil, & Van de Walle, 1995). These expectations guide object segmentati”
Learning Features by Watching Objects Move (2016, background): “Both infants [42] and newly sighted congenitally blind people [32] tend to oversegment static objects, but can group things properly when they move (Figure 1).”
Toward Causal Representation Learning (2021, background): “Infants’ understanding of physics relies upon objects that can be tracked over time and behave consistently [53], [236].”
Towards Causal Representation Learning (2021, background): “b) Issue 2 – Learning Reusable Mechanisms: Infants’ understanding of physics relies upon objects that can be tracked over time and behave consistently [52, 236].”
Flexible Neural Representation for Physics Prediction (2018, background): “Moreover, they either are born knowing, or quickly learn about, concepts such as object permanence, occlusion, and deformability, which guide their perception and reasoning [42].”
Relational Neural Expectation Maximization: Unsupervised Discovery of Objects and their Interactions (2018, background): “This demonstrates a notion of object permanence and aligns with evidence that young infants seem to infer that occluded objects move in connected paths and continue to maintain objectspecific properties (Spelke, 1990).”
A Compositional Object-Based Approach to Learning Physical Dynamics (2016, background): “For example, a foundational sense of intuitive physics is a prior that guides humans to decompose a scene into objects and carry expectations of object boundaries and motion across different scenarios (Spelke, 1990).”
Unmasking Clever Hans predictors and assessing what machines really learn (2019, background): “A major difference between machine and human learning is that humans (even infants) are equipped with a “start-up” software [62, 63] consisting of a priori representation about objects and physics [64,65] and about other agents [66,67].”
Automatically Composing Representation Transformations as a Means for Generalization (2018, background): “…cognitive development: abstract structural priors serve as a scaffolding within which concrete, domain-speciﬁc learning takes place [63, 78], but domain-speciﬁc learning about the continuous semantics of the world can also provide feedback to update the more discrete structural priors…”
Jointly Discovering Visual Objects and Spoken Words from Raw Sensory Input (2018, background): “stand speech and recognize objects in an extremely weakly supervised fashion, aided not by ground-truth annotations, but by observation, repetition, multi-modal context, and environmental interaction [1,2]. In this paper, we do not attempt to model the cognitive development of humans, but instead as”
The Theoretical Foundations of Evolutionary Psychology (2015, background): “…systems for these two domains: a mind-reading system for inferring the mental states of people (which can be selectively impaired in autism; Baron-Cohen, 1995; Leslie & Thaiss, 1992) and an object mechanics system for understanding the interactions of inanimate objects (Leslie, 1994; Spelke, 1990).”
Concepts and categories: Memory, meaning, and metaphysics (2012, background): “Learning in a given domain may be guided by certain skeletal principles, constraints, and (possibly innate) assumptions about the world (see Gelman & Coley, 1990; Gelman, 2003; Keil, 1981; Kellman & Spelke, 1983; Markman, 1990; Spelke, 1990).”
Infants Attribute Value to the Goal-Directed Actions of Self-propelled Objects (1997, background): “Shown objects that deviate from these properties, infants will increase their looking time (Spelke, 1990).”
Object Segmentation by Long Term Analysis of Point Trajectories (2010, background): “Biological vision systems learn objects up to a certain degree of accuracy in an unsupervised way by making use of the natural ordering of the images they see [1].”
Climate, Culture, and the Evolution of Cognition (2000, background): “Sophisticated sense systems depend upon powerful, specialized innate algorithms to make useful information from a mass of raw data from the sensory transducers (Shettleworth, this volume, Spelke, 1990).”
Conceptual precursors to language (2004, background): “Infants' sensitivity to this distinction is linked to representations of object mechanic”
Infants selectively encode the goal object of an actor's reach. (1998, background): “By 3 or 4 months of age, infants expect that two objects will not occupy the same space and that objects will exist and move continuously in time and space (Baillargeon, 1987, 1993; Spelke, 1990; Spelke et al., 1992).”