Resolving human object recognition in space and time

By Radoslaw Martin Cichy, Dimitrios Pantazis, Aude Oliva. Nature Neuroscience, 2014

837 citations recorded by the source corpus.

A comprehensive picture of object processing in the human brain requires combining both spatial and temporal information about brain activity. Here we acquired human magnetoencephalography (MEG) and functional magnetic resonance imaging (fMRI) responses to 92 object images. Multivariate pattern classification applied to MEG revealed the time course of object processing: whereas individual images were discriminated by visual representations early, ordinate and superordinate category levels emerged relatively late. Using representational similarity analysis, we combined human fMRI and MEG to show content-specific correspondence between early MEG responses and primary visual cortex (V1), and later MEG responses and inferior temporal (IT) cortex. We identified transient and persistent neural activities during object processing with sources in V1 and IT. Finally, we correlated human MEG signals to single-unit responses in monkey IT. Together, our findings provide an integrated space- and time-resolved view of human object categorization during the first few hundred milliseconds of vision.

Subfields: Cognitive Neuroscience

View Paper: https://www.semanticscholar.org/paper/6a50b5d18213300d2879453c3ccb577993682547

How researchers see this paper

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

In Cognitive Neuroscience

Citations in the corpus

Recurrence is required to capture the representational dynamics of the human visual system (2019, positive): “Results MEG data were recorded from 15 human participants (306 sensors, 2 sessions each) while they viewed 92 stimuli from a diverse set of natural object categories (human and nonhuman faces and bodies, natural and manmade inanimate objects) (16).”
Representation Similarity Analysis for Efficient Task Taxonomy & Transfer Learning (2019, methodology): “Motivated by the success of RSA in neuro-science [18, 4, 16, 1, 5, 25, 11], we investigate the application of RSA in obtaining task similarities (Figure 1b) and in transfer learning (Figure 1a).”
Comparison of deep neural networks to spatio-temporal cortical dynamics of human visual object recognition reveals hierarchical correspondence (2016, methodology): “The trained SVM was then used to predict the condition labels of the left-out testing data set consisting of the L th averaged pattern vector.”
Deep neural networks: a new framework for modelling biological vision and brain information processing (2015, background): “…automatic rapid interpretation of even static visual images (Sugase et al. 1999, Brincat & Connor 2006, Freiwald & Tsao 2010, Carlson et al. 2013, Cichy et al. 2014; Tang et al. 2014a), leading to the emergence of representations that clearly distinguish particular objects and object categories.”
Deep Supervised, but Not Unsupervised, Models May Explain IT Cortical Representation (2014, background): “Indeed, recent human magnetoencephalography studies using the same stimulus set [11,12] suggest that the major categorical divisions take slightly longer to emerge than a purely feedforward account would predict.”
Look and Think Twice: Capturing Top-Down Visual Attention with Feedback Convolutional Neural Networks (2015, background): “Despite great successes achieved by applying Feedforward Networks to image recognition and detection, evidences accumulate from cognitive studies and point to the feedback mechanism that may dominant human perception processes [3, 22, 16, 19].”
Neuroscience-Inspired Artificial Intelligence. (2017, methodology): “Machine learning techniques have transformed the analysis of neuroimaging datasets—for example, in the multivariate analysis of fMRI and magnetoencephalographic (MEG) data (Cichy et al., 2014; Çukur et al., 2013; Kriegeskorte and Kievit, 2013)—with promise for expediting connectomic analysis (Glasse”
Representational models: A common framework for understanding encoding, pattern-component, and representational-similarity analysis (2017, methodology): “16 using a sliding window over the time course of the trial [37-39], or by “stacking” the time series and 314”
A Primer on Pattern-Based Approaches to fMRI: Principles, Pitfalls, and Perspectives. (2015, background): “RSA also provides an important future direction for comparing coding spaces across different brain regions, or even between different modalities (Cichy et al., 2014) and different species”
A statistical framework for neuroimaging data analysis based on mutual information estimated via a gaussian copula (2016, background): “The common scale allows direct comparison of the strength of the modulation between different neuroimaging responses, and II opens up the promising possibility of directly relating the information content in different signals [Cichy et al., 2014].”
Dynamic Computational Time for Visual Attention (2017, methodology): “Inspired by human attention, the Recurrent Visual Attention Model (RAM) is proposed for image recognition [6].”
Decoding Dynamic Brain Patterns from Evoked Responses: A Tutorial on Multivariate Pattern Analysis Applied to Time Series Neuroimaging Data (2016, methodology): “Another popular alternative is the permutation test, which entails repeatedly shuffling the data and recomputing classifier performance on the shuffled data to obtain a null distribution, which is then compared against observed classifier performance on the original set to assess statistical signifi”
CoSMoMVPA: Multi-Modal Multivariate Pattern Analysis of Neuroimaging Data in Matlab/GNU Octave (2016, neutral): “…technique for fMRI, there are few reports of its application to magneto-and electro-encephalography (M/EEG 1) data (but see for example Kauhanen et al., 2006, Perreau Guimaraes, Wong, Uy, Grosenick, & Suppes, 2007, Pistohl, Schulze-Bonhage, & Aertsen, 2011, Chan, Halgren, Marinkovic, &…”
Don't fear ‘fear conditioning’: Methodological considerations for the design and analysis of studies on human fear acquisition, extinction, and return of fear (2017, neutral): “…equation models (SEMs) (Kuhn et al., 2016), Bayesian hierarchical modelling (Onat and Büchel, 2015), combined probability tests (e.g. Fisher’s or Brown’s method Brown, 1975), representational similarity analysis (Cichy et al., 2014), as well as classification approaches (Hogervorst et al., 2014).”
Distributed theta networks support the control of working memory: Evidence from scalp and intracranial EEG (2025, unknown): “Similarity analysis directly compares spatial patterns of neural activity across modalities to identify shared functional organization (Kriegeskorte et al., 2008; Cichy et al., 2014).”