There is ample evidence to show that the processing of face information involves a distributed neural network of face-sensitive areas in the occipitotemporal cortex and beyond (Duchaine and Yovel, 2015; Haxby et al., 2000). Three bilateral face-selective areas are considered as the core face-processing system, defined in functional Magnetic Resonance Imaging (fMRI) studies as regions showing significantly higher response to faces than objects, which are Occipital Face Area (OFA) in the inferior occipital gyrus (Gauthier et al., 2000; Haxby et al., 1999), Fusiform Face Area (FFA) in the fusiform gyrus (Kanwisher et al., 1997; Grill-Spector et al., 2004) and a face-sensitive area in the posterior superior temporal sulcus (pSTS) (Hoffman and Haxby, 2000; Puce et al., 1998). Similarly, a number of so-called face patches have been identified in macaque monkeys along the superior temporal sulcus (Tsao et al., 2003; Tsao et al., 2006; Tsao et al., 2008). Although the functional properties of these areas have been studied extensively, we do not yet have a comprehensive understanding of how the face-processing network functions in a dynamic manner. Hierarchical models postulate that face specific processes are initiated in the OFA based on local facial features, then the information is forwarded to higher level regions, such as FFA, for holistic processing (Haxby et al., 2000; Fairhall and Ishai, 2007; Liu et al., 2002). This model is supported by neuroimaging studies showing functional properties of face-selective areas and is consistent with generic local-to-global views of object processing. However, it has been challenged by results from studies in which patients with damaged OFA can still showed FFA activation to faces (Rossion et al., 2003; Steeves et al., 2006). Further, it was reported that during the perception of faces with minimal local facial features, FFA could still show face-preferential activation without face-selective inputs from OFA (Rossion et al., 2011). Thus a non-hierarchical model was proposed postulating that face detection is initiated at the FFA followed by a fine analysis in the OFA (Rossion et al., 2011; Gentile et al., 2017). These competing models may reflect different modes of operation of the face network under different demands. To reconcile these models, a comprehensive dynamic picture of face processing under different conditions with more detailed temporal information is needed.

In the current study, we investigated the dynamics of face processing in the ‘core face processing system’ using Magnetoencephalography (MEG) and fMRI. We designed the face-related stimuli specifically to reveal mechanisms for processing 1) normal faces, 2) Mooney faces with very little explicit facial features, 3) distorted faces with internal facial features spatially misarranged, and 4) contextually induced face representations with internal facial features completely missing. During the experiment, subjects were presented with various types of face pictures while MEG signals were recorded. The key effort in this study was in reconstructing the source signals from the MEG sensor data, to obtain a dynamic depiction of cortical responses to faces and other types of stimuli. With the timing of activation revealed in each face-selective area in the ‘core face processing system’, we could uncover when and where face information is processed in the human brain.

The main findings are briefly summarized here. First, we revealed the basic, mainly bottom-up, processing sequence along ventral temporal cortex by presenting face pictures of famous individuals to subjects. Face processing was initiated in the posterior areas and then proceeded forward to anterior regions. Right OFA (rOFA) and right posterior FFA (rpFFA) were activated very close in time, peaking around 120 ms, while right anterior FFA (raFFA) reached its peak at about 150 ms. The right pSTS (rpSTS) in the dorsal pathway showed a weaker and temporally more variable response, participating in face processing within a time window from 130 to 180 ms. Then, we highlighted the top-down operation in face processing by using two-tone Mooney face images (Mooney, 1957) lacking prototypical local facial features. According to the predictive coding theory (Rao and Ballard, 1999; Murray et al., 2004; Mumford, 1992), face prediction created at FFA based on impoverished information of Mooney faces and prior knowledge is poorly matched with the input representation at OFA due to the lack of explicit local facial features. The activity in OFA, representing ‘residual error’ between top-down prediction and bottom-up input, is then expected to increase subsequently. Consistent with this model, rOFA was activated later than rpFFA, and rpFFA exerted extensive directional influence onto rOFA when processing Mooney faces, suggesting a cortical analysis dominated by rpFFA to rOFA projection. However, when explicit internal facial features were available but misarranged within a normal face contour, a temporal pattern similar to that of normal faces was observed. Finally, we further investigated the temporal dynamics when face-specific responses were driven by contextual cues alone with the internal face features entirely missing (Cox et al., 2004). In this case, rOFA, rpFFA and raFFA were activated somewhat late and almost simultaneously, corresponding to contextual modulation that parallelly facilitated the processing of the core face-processing network.

Using a combined fMRI and MEG source localization approach, our results systematically revealed an intricately detailed dynamic picture of face information processing. Within the ventral occipitotemporal face processing network, normal faces were processed mainly in a bottom-up manner through the hierarchical pathway where input information was processed sequentially from posterior to anterior ventral temporal cortex. This temporal order was also observed when processing face-like stimuli with misarranged internal facial features. In contrast, during the processing of Mooney faces in the absence of prototypical facial features, top-down modulation was more prominent in which the dominant information flow was from the rpFFA to rOFA. Moreover, face-specific responses from contextual cues alone were evoked late and simultaneously across the rOFA, rpFFA and raFFA, suggesting that contextual facilitation acted parallelly on the core face-processing network. These results advance our understanding of the hierarchical and non-hierarchical models of face perception, especially underscoring the stimulus- and context-dependent nature of the processing sequences.

During the perception of 2-tone Mooney faces, it is necessary to discount shadows and recover 3D surface structure from 2D images (Grützner et al., 2010). Interestingly, only familiar objects, like faces, can be interpreted to be volumetric easily from 2-tone representations (Moore and Cavanagh, 1998; Hegde et al., 2007). Thus it is supposed that prior knowledge should play an important role in the recovery of 3D shape from Mooney images (Braje et al., 1998; Gerardin et al., 2010). A top-down model emphasized the guidance of prior experience at higher levels (Cavanagh, 1991). This model is supported by evidence from experiments showing that early visual processing is affected by high-level attributes in both human and monkey (Lee et al., 2002; Humphrey et al., 1997; Issa et al., 2018). As briefly mentioned in the results section, the dynamics of MEG signals associated with processing Mooney faces, which highlights the top-down modulation, is consistent with the explanation based on predictive coding model. It proposed that hypotheses or predictions made at higher cortical areas are compared with, through feedback, representations at lower areas to generate residual error, which is then forwarded to higher stages as ‘neural activity’ (Rao and Ballard, 1999; Murray et al., 2004; Friston, 2005; Friston, 2010). Specifically, the face model/prediction is generated at the rpFFA based on the global configuration of Mooney faces using prior knowledge about 3D faces, illumination, and cast shadows. This prediction of expected facial features is then poorly matched with the input representation at the rOFA which lacks the explicit prototypical facial features due to the mixed illumination-invariant and illumination-dependent features, generating an increased signal at rOFA. Thus, the dominant signal at the rOFA (residual) necessarily lags behind the signal at the rpFFA (hypothesis). However, when processing normal faces or face with misarranged facial features, the prominent signal in the early stage of rOFA is mainly due to the strong feedforward input from early visual cortex as rOFA is robustly responsive to the clear facial components. The prediction feedback from rpFFA would be consistent with representation at the rOFA in the case of the normal faces, resulting in little error signal; with the misarranged facial features, there was a hint of a late increase of rOFA signal, possibly indicating that the feedback signal could contain some spatial information as well.

The timing of face induced neural activation has been studied for a long time with various techniques, such as the combination of MEG and fMRI using representational similarities (Cichy et al., 2014; Cichy et al., 2016), MEG source localization and intracranial EEG (Kadipasaoglu et al., 2017; Keller et al., 2017; Ghuman et al., 2014). An early MEG study suggested two stages (early categorization and late identification) were involved in face processing (Liu et al., 2002). Combined with the fMRI observation that OFA is responsible for identifying facial parts while FFA for holistic configuration (Rotshtein et al., 2005; Liu et al., 2010; Pitcher et al., 2011b; Arcurio et al., 2012; Pitcher et al., 2007; Schiltz, 2010), OFA is expected to respond earlier than FFA. An simultaneous electroencephalogram (EEG)-fMRI study also showed that OFA responded to faces earlier than FFA (OFA: 110 ms; FFA: 170 ms) (Sadeh et al., 2010). Using transient stimulation to temporally disrupt local neural processing, Transcranial Magnetic Stimulation (TMS) experiments suggested that OFA processes facial information at about 100/110 ms, while pSTS begins processing face at about 100/140 ms (Pitcher et al., 2012; Pitcher et al., 2014). However, the sources of N/M170 face selective component remain controversial, it is suggested to come from fusiform gyrus in some studies (Deffke et al., 2007; Kume et al., 2016; Perry and Singh, 2014). While some other studies emphasized the contribution of inferior occipital gyrus besides fusiform gyrus (Itier et al., 2006; Gao et al., 2013) or even of pSTS (Nguyen and Cunnington, 2014). Our results provide more precise and detailed timing information of the core face network under various stimulus and contextual conditions, especially the temporal relationship between rpFFA and raFFA. raFFA is engaged significantly later, about 20 ms after the rpFFA, suggesting that the raFFA likely plays a different functional role from rpFFA. This idea is supported by previous anatomical evidence showing that pFFA and aFFA have different cellular architectures (Weiner et al., 2017).

Our results also shed light on the role of internal and external features in face perception. Although when assembling into a whole face, facial features are processed holistically and the representation of internal features are influenced by external features (Andrews et al., 2010), eyes in isolation elicit a later but larger N170 (Bentin et al., 1996; Rossion and Jacques, 2011) and can drive face-selective neurons as well as full-face images (Issa and DiCarlo, 2012) in monkeys. In our results, the somewhat slower but still sequential progression of face responses elicited by face-like stimuli with clear but misarranged internal features in face outline further supports that facial features are sufficient to trigger the bottom-up face processing sequence. In addition, certain stimulus manipulations, such as face inversion (Bentin et al., 1996), contrast reversal, Mooney transformation or removal of facial features produced comparable (or even increased amplitude) but delayed N170 responses (Rossion and Jacques, 2011). Thus it is suggested that as long as the impoverished stimuli is perceived as a face, inferior temporal cortex areas would be activated (McKeeff and Tong, 2007; Grützner et al., 2010). Our results provide further more details for this explanation by showing the top-down rpFFA to rOFA projection when the prototypical facial features are lack.

Besides facial features, contextual information is also important for face interpretation (Chen and Whitney, 2019; Martinez, 2019). Interestingly, FFA can be activated by the perceived presence of faces from contextual body cues alone (Cox et al., 2004). Here our MEG data showed that the face-selective areas in ventral core face network were indeed activated by the contextual cues for faces, but they were not activated in any order, instead, they became active together at a late stage. This is similar to the temporal dynamics observed in visual imagery, a top-down process given the absence of visual inputs (Dijkstra et al., 2018). Future studies are needed to elucidate how core face network interacts with other brain regions to trigger the face perception. For example, according to a MEG study using fast periodic visual stimulation approach (Rossion et al., 2012; Rossion et al., 2015; de Heering and Rossion, 2015), top-down attention increase the response in FFA by gamma synchrony between the inferior frontal junction and FFA (Baldauf and Desimone, 2014).

Face perception is shaped by long-term visual experience, for example, familiar faces are processed more efficiently than unfamiliar ones (Landi and Freiwald, 2017; Schwartz and Yovel, 2016; Dobs et al., 2019; Gobbini and Haxby, 2006). In terms of the dynamics in the ventral occipitotemporal areas, the present results showed little differences between processing famous and unfamiliar faces. This could be due to several reasons. First, many studies suggested that regions in the anterior temporal lobe rather than OFA and FFA represent face familiarity (Gobbini and Haxby, 2007; Pourtois et al., 2005; Sugiura et al., 2011). However, extended face system is beyond the scope of our current study because some areas in the extended system are too deep to obtain a good MEG source signal. Second, some subjects might not be familiar with all the famous faces we used. Third, familiarity may affect face recognition via high gamma frequency band activity (Anaki et al., 2007), which is not included in our data analysis.

Bilateral pSTS showed weak and multi-peaked responses during both famous and unfamiliar face processing despite the task differences. One possible reason for the multiple peaks of responses is that as a hub for integrating information from multiple sources (e.g., face, body, and voice), STS contains regions that respond to different types of information (Grossman et al., 2005; Bernstein and Yovel, 2015). A lot of studies have suggested diverse functional role of pSTS in representing changeable aspects of faces, such as expression, lip movement and eye-gaze (Baseler et al., 2014; Engell and Haxby, 2007). Specifically, pSTS is involved in the analysis of facial muscle articulations which are combined to produce facial expressions (Srinivasan et al., 2016; Martinez, 2017). In addition, pSTS may respond to dynamic motion information conveyed through faces (O'Toole et al., 2002).

Previous studies showed that left and right fusiform gyrus are differentially involved in face/non-face judgements (Meng et al., 2012; Goold and Meng, 2017), ‘low-level’ face semblance and perceptual learning of face (Bi et al., 2014; Feng et al., 2011; McGugin et al., 2018). Interestingly, in our results, the peak latency of the left pFFA was later than that of the right pFFA in all conditions except famous face. Responses evoked from distorted faces with misarranged features had the largest lateral difference (20 ms). One possible reason is that the signal attributed to the left pFFA is in fact a mixture of signals from pFFA and aFFA.

Although the exact correspondence between human and macaque face-selective areas are still unclear (Tsao et al., 2003; Tsao et al., 2006; Tsao et al., 2008), the dynamic picture of normal face processing revealed in our study is generally similar to that in macaques. Single-unit recording studies showed that activity begins slightly earlier in posterior face patches than anterior ones, reaching peak levels around 126, 133, and 145 ms for middle lateral (ML)/middle fundus (MF), anterior lateral (AL), and anterior medial (AM) (Freiwald and Tsao, 2010) , respectively. Interestingly, there is a discrepancy in response to Mooney faces in high level face patch AM between two monkeys. One of them showed nearly the same peak latency as normal faces but with more sustained activation, while the other did not response to Mooney faces (Moeller et al., 2017). This may imply that the processing of Mooney faces is related to individual face detection ability or life experience and face processing is not a simple feedforward process from low level to high level areas. Consistent with that, a more recent study showed a rapid and more sustained response in high level face area (aIT) and an early rising then quickly decreased activity in low level areas in monkeys, a signature of predictive coding model (Issa et al., 2018).

Our study is obviously limited in scope. There are many types of cues and tasks relevant for face perception that could be investigated. In addition to facial features and context, many low level cues contribute to face recognition, such as illumination direction, pigmentation (surface appearance) and contrast polarity (one region brighter than another) (Russell et al., 2007; Sinha et al., 2006). In particular, neurons tuned for contrast polarity were found in macaque inferotemporal cortex, supporting the notion that low-level image properties are encoded in face regions (Ohayon et al., 2012; Weibert et al., 2018). We purposely avoided the complication of color cues in this study by using gray-scale images, but we are aware the importance of color in face perception (Yip and Sinha, 2002; Benitez-Quiroz et al., 2018). Moreover, the temporal dynamics of face processing could very well be influenced by different tasks. In our results, there is little difference between the temporal patterns in response to unfamiliar faces under face category task (Figure 2—figure supplement 1) and image identity one-back task (Figure 3). Future studies are needed to more comprehensively investigate the role of behavioral tasks, especially during the relatively late stages of face processing.

In summary, our study delineated the precise timing of bottom-up, top-down, as well as context-facilitated processing sequences in the occipital-temporal face network. These results provide a way to understand and reconcile previous discrepant findings, revealing the dominant bottom-up processing when explicit facial features were present, and highlighting the importance of the top-down feedback operations when faced with impoverished inputs with unclear or ambiguous facial features.

The source data files have been provided for Figures 2, 3, 4, 5 and Figure 2—figure supplement 1. MEG source activation data (processed based on original fMRI and MEG datasets ) have been deposited in Open Science Framework and can be accessed at https://osf.io/vhefz/.

1. 1. Fan X

   (2020) Open Science Framework

   MEG face experiments.

   https://osf.io/vhefz/

1. 1. Anaki D
   2. Zion-Golumbic E
   3. Bentin S

   (2007) Electrophysiological neural mechanisms for detection, configural analysis and recognition of faces

   NeuroImage 37:1407–1416.

   https://doi.org/10.1016/j.neuroimage.2007.05.054

   • PubMed
   • Google Scholar
2. 1. Andrews TJ
   2. Davies-Thompson J
   3. Kingstone A
   4. Young AW

   (2010) Internal and external features of the face are represented holistically in face-selective regions of visual cortex

   Journal of Neuroscience 30:3544–3552.

   https://doi.org/10.1523/JNEUROSCI.4863-09.2010

   • PubMed
   • Google Scholar
3. 1. Arcurio LR
   2. Gold JM
   3. James TW

   (2012) The response of face-selective cortex with single face parts and part combinations

   Neuropsychologia 50:2454–2459.

   https://doi.org/10.1016/j.neuropsychologia.2012.06.016

   • PubMed
   • Google Scholar
4. 1. Ashrafulla S
   2. Haldar JP
   3. Joshi AA
   4. Leahy RM

   (2013) Canonical Granger causality between regions of interest

   NeuroImage 83:189–199.

   https://doi.org/10.1016/j.neuroimage.2013.06.056

   • PubMed
   • Google Scholar
5. 1. Baldauf D
   2. Desimone R

   (2014) Neural mechanisms of object-based attention

   Science 344:424–427.

   https://doi.org/10.1126/science.1247003

   • PubMed
   • Google Scholar
6. 1. Barnett L
   2. Seth AK

   (2014) The MVGC multivariate granger causality toolbox: a new approach to Granger-causal inference

   Journal of Neuroscience Methods 223:50–68.

   https://doi.org/10.1016/j.jneumeth.2013.10.018

   • PubMed
   • Google Scholar
7. 1. Baseler HA
   2. Harris RJ
   3. Young AW
   4. Andrews TJ

   (2014) Neural responses to expression and gaze in the posterior superior temporal sulcus interact with facial identity

   Cerebral Cortex 24:737–744.

   https://doi.org/10.1093/cercor/bhs360

   • PubMed
   • Google Scholar
8. 1. Benitez-Quiroz CF
   2. Srinivasan R
   3. Martinez AM

   (2018) Facial color is an efficient mechanism to visually transmit emotion

   PNAS 115:3581–3586.

   https://doi.org/10.1073/pnas.1716084115

   • PubMed
   • Google Scholar
9. 1. Benjamini Y
   2. Hochberg Y

   (1995) Controlling the false discovery rate: a practical and powerful approach to multiple testing

   Journal of the Royal Statistical Society: Series B 57:289–300.

   https://doi.org/10.1111/j.2517-6161.1995.tb02031.x

   • Google Scholar
10. 1. Bentin S
    2. Allison T
    3. Puce A
    4. Perez E
    5. McCarthy G

    (1996) Electrophysiological studies of face perception in humans

    Journal of Cognitive Neuroscience 8:551–565.

    https://doi.org/10.1162/jocn.1996.8.6.551

    • PubMed
    • Google Scholar
11. 1. Bernstein M
    2. Yovel G

    (2015) Two neural pathways of face processing: a critical evaluation of current models

    Neuroscience & Biobehavioral Reviews 55:536–546.

    https://doi.org/10.1016/j.neubiorev.2015.06.010

    • PubMed
    • Google Scholar
12. 1. Bi T
    2. Chen J
    3. Zhou T
    4. He Y
    5. Fang F

    (2014) Function and structure of human left fusiform cortex are closely associated with perceptual learning of faces

    Current Biology 24:222–227.

    https://doi.org/10.1016/j.cub.2013.12.028

    • PubMed
    • Google Scholar
13. 1. Braje W
    2. Kersten D
    3. Tarr M
    4. Troje N

    (1998) Illumination effects in face recognition

    Psychobiology 26:371–380.

    https://doi.org/10.3758/BF03330623

    • Google Scholar
14. Book

    1. Cavanagh P

    (1991)

    What’s up in top-down processing

    In: Gorea A, editors. Representations of Vision. Cambridge University Press. pp. 295–304.

    • Google Scholar
15. 1. Chen Z
    2. Whitney D

    (2019) Tracking the affective state of unseen persons

    PNAS 116:7559–7564.

    https://doi.org/10.1073/pnas.1812250116

    • PubMed
    • Google Scholar
16. 1. Cichy RM
    2. Pantazis D
    3. Oliva A

    (2014) Resolving human object recognition in space and time

    Nature Neuroscience 17:455–462.

    https://doi.org/10.1038/nn.3635

    • PubMed
    • Google Scholar
17. 1. Cichy RM
    2. Pantazis D
    3. Oliva A

    (2016) Similarity-Based fusion of MEG and fMRI reveals Spatio-Temporal dynamics in human cortex during visual object recognition

    Cerebral Cortex 26:3563–3579.

    https://doi.org/10.1093/cercor/bhw135

    • PubMed
    • Google Scholar
18. 1. Cox D
    2. Meyers E
    3. Sinha P

    (2004) Contextually evoked object-specific responses in human visual cortex

    Science 304:115–117.

    https://doi.org/10.1126/science.1093110

    • PubMed
    • Google Scholar
19. 1. de Heering A
    2. Rossion B

    (2015) Rapid categorization of natural face images in the infant right hemisphere

    eLife 4:e06564.

    https://doi.org/10.7554/eLife.06564

    • PubMed
    • Google Scholar
20. 1. Deffke I
    2. Sander T
    3. Heidenreich J
    4. Sommer W
    5. Curio G
    6. Trahms L
    7. Lueschow A

    (2007) MEG/EEG sources of the 170-ms response to faces are co-localized in the fusiform gyrus

    NeuroImage 35:1495–1501.

    https://doi.org/10.1016/j.neuroimage.2007.01.034

    • PubMed
    • Google Scholar
21. 1. Dijkstra N
    2. Mostert P
    3. Lange FP
    4. Bosch S
    5. van Gerven MA

    (2018) Differential temporal dynamics during visual imagery and perception

    eLife 7:e33904.

    https://doi.org/10.7554/eLife.33904

    • PubMed
    • Google Scholar
22. 1. Dobs K
    2. Isik L
    3. Pantazis D
    4. Kanwisher N

    (2019) How face perception unfolds over time

    Nature Communications 10:1258.

    https://doi.org/10.1038/s41467-019-09239-1

    • PubMed
    • Google Scholar
23. 1. Duchaine B
    2. Yovel G

    (2015) A revised neural framework for face processing

    Annual Review of Vision Science 1:393–416.

    https://doi.org/10.1146/annurev-vision-082114-035518

    • PubMed
    • Google Scholar
24. 1. Engell AD
    2. Haxby JV

    (2007) Facial expression and gaze-direction in human superior temporal sulcus

    Neuropsychologia 45:3234–3241.

    https://doi.org/10.1016/j.neuropsychologia.2007.06.022

    • PubMed
    • Google Scholar
25. 1. Fairhall SL
    2. Ishai A

    (2007) Effective connectivity within the distributed cortical network for face perception

    Cerebral Cortex 17:2400–2406.

    https://doi.org/10.1093/cercor/bhl148

    • PubMed
    • Google Scholar
26. 1. Feng L
    2. Liu J
    3. Wang Z
    4. Li J
    5. Li L
    6. Ge L
    7. Tian J
    8. Lee K

    (2011) The other face of the other-race effect: an fMRI investigation of the other-race face categorization advantage

    Neuropsychologia 49:3739–3749.

    https://doi.org/10.1016/j.neuropsychologia.2011.09.031

    • PubMed
    • Google Scholar
27. 1. Freiwald WA
    2. Tsao DY

    (2010) Functional compartmentalization and viewpoint generalization within the macaque face-processing system

    Science 330:845–851.

    https://doi.org/10.1126/science.1194908

    • PubMed
    • Google Scholar
28. 1. Friston K

    (2005) A theory of cortical responses

    Philosophical Transactions of the Royal Society B: Biological Sciences 360:815–836.

    https://doi.org/10.1098/rstb.2005.1622

    • Google Scholar
29. 1. Friston K

    (2010) The free-energy principle: a unified brain theory?

    Nature Reviews Neuroscience 11:127–138.

    https://doi.org/10.1038/nrn2787

    • PubMed
    • Google Scholar
30. 1. Gao Z
    2. Goldstein A
    3. Harpaz Y
    4. Hansel M
    5. Zion-Golumbic E
    6. Bentin S

    (2013) A magnetoencephalographic study of face processing: m170, gamma-band oscillations and source localization

    Human Brain Mapping 34:1783–1795.

    https://doi.org/10.1002/hbm.22028

    • PubMed
    • Google Scholar
31. 1. Gauthier I
    2. Tarr MJ
    3. Moylan J
    4. Skudlarski P
    5. Gore JC
    6. Anderson AW

    (2000) The fusiform "face area" is part of a network that processes faces at the individual level

    Journal of Cognitive Neuroscience 12:495–504.

    https://doi.org/10.1162/089892900562165

    • PubMed
    • Google Scholar
32. 1. Gentile F
    2. Ales J
    3. Rossion B

    (2017) Being BOLD: the neural dynamics of face perception

    Human Brain Mapping 38:120–139.

    https://doi.org/10.1002/hbm.23348

    • PubMed
    • Google Scholar
33. 1. Gerardin P
    2. Kourtzi Z
    3. Mamassian P

    (2010) Prior knowledge of illumination for 3D perception in the human brain

    PNAS 107:16309–16314.

    https://doi.org/10.1073/pnas.1006285107

    • PubMed
    • Google Scholar
34. 1. Ghuman AS
    2. Brunet NM
    3. Li Y
    4. Konecky RO
    5. Pyles JA
    6. Walls SA
    7. Destefino V
    8. Wang W
    9. Richardson RM

    (2014) Dynamic encoding of face information in the human fusiform gyrus

    Nature Communications 5:5672.

    https://doi.org/10.1038/ncomms6672

    • PubMed
    • Google Scholar
35. 1. Gobbini MI
    2. Haxby JV

    (2006) Neural response to the visual familiarity of faces

    Brain Research Bulletin 71:76–82.

    https://doi.org/10.1016/j.brainresbull.2006.08.003

    • PubMed
    • Google Scholar
36. 1. Gobbini MI
    2. Haxby JV

    (2007) Neural systems for recognition of familiar faces

    Neuropsychologia 45:32–41.

    https://doi.org/10.1016/j.neuropsychologia.2006.04.015

    • PubMed
    • Google Scholar
37. 1. Goold JE
    2. Meng M

    (2017) Categorical learning revealed in activity pattern of left fusiform cortex

    Human Brain Mapping 38:3648–3658.

    https://doi.org/10.1002/hbm.23620

    • PubMed
    • Google Scholar
38. 1. Gramfort A
    2. Luessi M
    3. Larson E
    4. Engemann DA
    5. Strohmeier D
    6. Brodbeck C
    7. Goj R
    8. Jas M
    9. Brooks T
    10. Parkkonen L
    11. Hämäläinen M

    (2013) MEG and EEG data analysis with MNE-Python

    Frontiers in Neuroscience 7:267.

    https://doi.org/10.3389/fnins.2013.00267

    • PubMed
    • Google Scholar
39. 1. Gramfort A
    2. Luessi M
    3. Larson E
    4. Engemann DA
    5. Strohmeier D
    6. Brodbeck C
    7. Parkkonen L
    8. Hämäläinen MS

    (2014) MNE software for processing MEG and EEG data

    NeuroImage 86:446–460.

    https://doi.org/10.1016/j.neuroimage.2013.10.027

    • PubMed
    • Google Scholar
40. 1. Granger CWJ

    (1969) Investigating causal relations by econometric models and Cross-spectral methods

    Econometrica 37:424.

    https://doi.org/10.2307/1912791

    • Google Scholar
41. 1. Grill-Spector K
    2. Knouf N
    3. Kanwisher N

    (2004) The fusiform face area subserves face perception, not generic within-category identification

    Nature Neuroscience 7:555–562.

    https://doi.org/10.1038/nn1224

    • PubMed
    • Google Scholar
42. 1. Grossman ED
    2. Battelli L
    3. Pascual-Leone A

    (2005) Repetitive TMS over posterior STS disrupts perception of biological motion

    Vision Research 45:2847–2853.

    https://doi.org/10.1016/j.visres.2005.05.027

    • Google Scholar
43. 1. Grützner C
    2. Uhlhaas PJ
    3. Genc E
    4. Kohler A
    5. Singer W
    6. Wibral M

    (2010) Neuroelectromagnetic correlates of perceptual closure processes

    Journal of Neuroscience 30:8342–8352.

    https://doi.org/10.1523/JNEUROSCI.5434-09.2010

    • PubMed
    • Google Scholar
44. 1. Haxby JV
    2. Ungerleider LG
    3. Clark VP
    4. Schouten JL
    5. Hoffman EA
    6. Martin A

    (1999) The effect of face inversion on activity in human neural systems for face and object perception

    Neuron 22:189–199.

    https://doi.org/10.1016/S0896-6273(00)80690-X

    • Google Scholar
45. 1. Haxby JV
    2. Hoffman EA
    3. Gobbini MI

    (2000) The distributed human neural system for face perception

    Trends in Cognitive Sciences 4:223–233.

    https://doi.org/10.1016/S1364-6613(00)01482-0

    • PubMed
    • Google Scholar
46. 1. Hegde J
    2. Thompson S
    3. Kersten D

    (2007) Identifying faces in two-tone ('Mooney') images: A psychophysical and fMRI study

    Journal of Vision 7:624.

    https://doi.org/10.1167/7.9.624

    • Google Scholar
47. 1. Hoffman EA
    2. Haxby JV

    (2000) Distinct representations of eye gaze and identity in the distributed human neural system for face perception

    Nature Neuroscience 3:80–84.

    https://doi.org/10.1038/71152

    • PubMed
    • Google Scholar
48. 1. Humphrey GK
    2. Goodale MA
    3. Bowen CV
    4. Gati JS
    5. Vilis T
    6. Rutt BK
    7. Menon RS

    (1997) Differences in perceived shape from shading correlate with activity in early visual Areas

    Current Biology 7:144–147.

    https://doi.org/10.1016/S0960-9822(06)00058-3

    • PubMed
    • Google Scholar
49. 1. Issa EB
    2. Cadieu CF
    3. DiCarlo JJ

    (2018) Neural dynamics at successive stages of the ventral visual stream are consistent with hierarchical error signals

    eLife 7:e42870.

    https://doi.org/10.7554/eLife.42870

    • PubMed
    • Google Scholar
50. 1. Issa EB
    2. DiCarlo JJ

    (2012) Precedence of the eye region in neural processing of faces

    Journal of Neuroscience 32:16666–16682.

    https://doi.org/10.1523/JNEUROSCI.2391-12.2012

    • PubMed
    • Google Scholar
51. 1. Itier RJ
    2. Herdman AT
    3. George N
    4. Cheyne D
    5. Taylor MJ

    (2006) Inversion and contrast-reversal effects on face processing assessed by MEG

    Brain Research 1115:108–120.

    https://doi.org/10.1016/j.brainres.2006.07.072

    • PubMed
    • Google Scholar
52. 1. Kadipasaoglu CM
    2. Conner CR
    3. Baboyan VG
    4. Rollo M
    5. Pieters TA
    6. Tandon N

    (2017) Network dynamics of human face perception

    PLOS ONE 12:e0188834.

    https://doi.org/10.1371/journal.pone.0188834

    • PubMed
    • Google Scholar
53. 1. Kanwisher N
    2. McDermott J
    3. Chun MM

    (1997) The fusiform face area: a module in human extrastriate cortex specialized for face perception

    The Journal of Neuroscience 17:4302–4311.

    https://doi.org/10.1523/JNEUROSCI.17-11-04302.1997

    • PubMed
    • Google Scholar
54. 1. Keller CJ
    2. Davidesco I
    3. Megevand P
    4. Lado FA
    5. Malach R
    6. Mehta AD

    (2017) Tuning face perception with electrical stimulation of the fusiform gyrus

    Human Brain Mapping 38:2830–2842.

    https://doi.org/10.1002/hbm.23543

    • PubMed
    • Google Scholar
55. 1. Kume Y
    2. Maekawa T
    3. Urakawa T
    4. Hironaga N
    5. Ogata K
    6. Shigyo M
    7. Tobimatsu S

    (2016) Neuromagnetic evidence that the right fusiform face area is essential for human face awareness: an intermittent binocular rivalry study

    Neuroscience Research 109:54–62.

    https://doi.org/10.1016/j.neures.2016.02.004

    • PubMed
    • Google Scholar
56. 1. Landi SM
    2. Freiwald WA

    (2017) Two Areas for familiar face recognition in the primate brain

    Science 357:591–595.

    https://doi.org/10.1126/science.aan1139

    • PubMed
    • Google Scholar
57. 1. Lee TS
    2. Yang CF
    3. Romero RD
    4. Mumford D

    (2002) Neural activity in early visual cortex reflects behavioral experience and higher-order perceptual saliency

    Nature Neuroscience 5:589–597.

    https://doi.org/10.1038/nn0602-860

    • PubMed
    • Google Scholar
58. 1. Liu J
    2. Harris A
    3. Kanwisher N

    (2002) Stages of processing in face perception: an MEG study

    Nature Neuroscience 5:910–916.

    https://doi.org/10.1038/nn909

    • PubMed
    • Google Scholar
59. 1. Liu J
    2. Harris A
    3. Kanwisher N

    (2010) Perception of face parts and face configurations: an FMRI study

    Journal of Cognitive Neuroscience 22:203–211.

    https://doi.org/10.1162/jocn.2009.21203

    • PubMed
    • Google Scholar
60. 1. Maris E
    2. Oostenveld R

    (2007) Nonparametric statistical testing of EEG- and MEG-data

    Journal of Neuroscience Methods 164:177–190.

    https://doi.org/10.1016/j.jneumeth.2007.03.024

    • PubMed
    • Google Scholar
61. 1. Martinez AM

    (2017) Visual perception of facial expressions of emotion

    Current Opinion in Psychology 17:27–33.

    https://doi.org/10.1016/j.copsyc.2017.06.009

    • PubMed
    • Google Scholar
62. 1. Martinez AM

    (2019) Context may reveal how you feel

    PNAS 116:7169–7171.

    https://doi.org/10.1073/pnas.1902661116

    • PubMed
    • Google Scholar
63. 1. McGugin RW
    2. Ryan KF
    3. Tamber-Rosenau BJ
    4. Gauthier I

    (2018) The role of experience in the Face-Selective response in right FFA

    Cerebral Cortex 28:2071–2084.

    https://doi.org/10.1093/cercor/bhx113

    • PubMed
    • Google Scholar
64. 1. McKeeff TJ
    2. Tong F

    (2007) The timing of perceptual decisions for ambiguous face stimuli in the human ventral visual cortex

    Cerebral Cortex 17:669–678.

    https://doi.org/10.1093/cercor/bhk015

    • PubMed
    • Google Scholar
65. 1. Meng M
    2. Cherian T
    3. Singal G
    4. Sinha P

    (2012) Lateralization of face processing in the human brain

    Proceedings of the Royal Society B: Biological Sciences 279:2052–2061.

    https://doi.org/10.1098/rspb.2011.1784

    • Google Scholar
66. 1. Moeller S
    2. Crapse T
    3. Chang L
    4. Tsao DY

    (2017) The effect of face patch microstimulation on perception of faces and objects

    Nature Neuroscience 20:743–752.

    https://doi.org/10.1038/nn.4527

    • PubMed
    • Google Scholar
67. 1. Mooney CM

    (1957) Age in the development of closure ability in children

    Canadian Journal of Psychology/Revue Canadienne De Psychologie 11:219–226.

    https://doi.org/10.1037/h0083717

    • Google Scholar
68. 1. Moore C
    2. Cavanagh P

    (1998) Recovery of 3D volume from 2-tone images of novel objects

    Cognition 67:45–71.

    https://doi.org/10.1016/S0010-0277(98)00014-6

    • PubMed
    • Google Scholar
69. 1. Mumford D

    (1992) On the computational architecture of the neocortex. II. the role of cortico-cortical loops

    Biological Cybernetics 66:241–251.

    https://doi.org/10.1007/bf00198477

    • PubMed
    • Google Scholar
70. 1. Murray SO
    2. Schrater P
    3. Kersten D

    (2004) Perceptual grouping and the interactions between visual cortical Areas

    Neural Networks 17:695–705.

    https://doi.org/10.1016/j.neunet.2004.03.010

    • PubMed
    • Google Scholar
71. 1. Nguyen VT
    2. Cunnington R

    (2014) The superior temporal sulcus and the N170 during face processing: single trial analysis of concurrent EEG-fMRI

    NeuroImage 86:492–502.

    https://doi.org/10.1016/j.neuroimage.2013.10.047

    • PubMed
    • Google Scholar
72. 1. O'Toole AJ
    2. Roark DA
    3. Abdi H

    (2002) Recognizing moving faces: a psychological and neural synthesis

    Trends in Cognitive Sciences 6:261–266.

    https://doi.org/10.1016/S1364-6613(02)01908-3

    • PubMed
    • Google Scholar
73. 1. Ohayon S
    2. Freiwald WA
    3. Tsao DY

    (2012) What makes a cell face selective? the importance of contrast

    Neuron 74:567–581.

    https://doi.org/10.1016/j.neuron.2012.03.024

    • PubMed
    • Google Scholar
74. 1. Oostenveld R
    2. Fries P
    3. Maris E
    4. Schoffelen JM

    (2011) FieldTrip: open source software for advanced analysis of MEG, EEG, and invasive electrophysiological data

    Computational Intelligence and Neuroscience 2011:1–9.

    https://doi.org/10.1155/2011/156869

    • PubMed
    • Google Scholar
75. 1. Perry G
    2. Singh KD

    (2014) Localizing evoked and induced responses to faces using magnetoencephalography

    European Journal of Neuroscience 39:1517–1527.

    https://doi.org/10.1111/ejn.12520

    • PubMed
    • Google Scholar
76. 1. Pitcher D
    2. Walsh V
    3. Yovel G
    4. Duchaine B

    (2007) TMS evidence for the involvement of the right occipital face area in early face processing

    Current Biology 17:1568–1573.

    https://doi.org/10.1016/j.cub.2007.07.063

    • PubMed
    • Google Scholar
77. 1. Pitcher D
    2. Dilks DD
    3. Saxe RR
    4. Triantafyllou C
    5. Kanwisher N

    (2011a) Differential selectivity for dynamic versus static information in face-selective cortical regions

    NeuroImage 56:2356–2363.

    https://doi.org/10.1016/j.neuroimage.2011.03.067

    • PubMed
    • Google Scholar
78. 1. Pitcher D
    2. Walsh V
    3. Duchaine B

    (2011b) The role of the occipital face area in the cortical face perception network

    Experimental Brain Research 209:481–493.

    https://doi.org/10.1007/s00221-011-2579-1

    • PubMed
    • Google Scholar
79. 1. Pitcher D
    2. Goldhaber T
    3. Duchaine B
    4. Walsh V
    5. Kanwisher N

    (2012) Two critical and functionally distinct stages of face and body perception

    Journal of Neuroscience 32:15877–15885.

    https://doi.org/10.1523/JNEUROSCI.2624-12.2012

    • PubMed
    • Google Scholar
80. 1. Pitcher D
    2. Duchaine B
    3. Walsh V

    (2014) Combined TMS and FMRI reveal dissociable cortical pathways for dynamic and static face perception

    Current Biology 24:2066–2070.

    https://doi.org/10.1016/j.cub.2014.07.060

    • PubMed
    • Google Scholar
81. 1. Pourtois G
    2. Schwartz S
    3. Seghier ML
    4. Lazeyras F
    5. Vuilleumier P

    (2005) View-independent coding of face identity in frontal and temporal cortices is modulated by familiarity: an event-related fMRI study

    NeuroImage 24:1214–1224.

    https://doi.org/10.1016/j.neuroimage.2004.10.038

    • PubMed
    • Google Scholar
82. 1. Puce A
    2. Allison T
    3. Bentin S
    4. Gore JC
    5. McCarthy G

    (1998) Temporal cortex activation in humans viewing eye and mouth movements

    The Journal of Neuroscience 18:2188–2199.

    https://doi.org/10.1523/JNEUROSCI.18-06-02188.1998

    • PubMed
    • Google Scholar
83. 1. Rao R
    2. Ballard DH

    (1999) Predictive coding in the visual cortex: a functional interpretation of some extra-classical receptive-field effects

    Nature 2:79–87.

    https://doi.org/10.1038/4580

    • Google Scholar
84. 1. Rossion B
    2. Caldara R
    3. Seghier M
    4. Schuller AM
    5. Lazeyras F
    6. Mayer E

    (2003) A network of occipito-temporal face-sensitive Areas besides the right middle fusiform gyrus is necessary for normal face processing

    Brain 126:2381–2395.

    https://doi.org/10.1093/brain/awg241

    • PubMed
    • Google Scholar
85. 1. Rossion B
    2. Dricot L
    3. Goebel R
    4. Busigny T

    (2011) Holistic face categorization in higher order visual Areas of the normal and prosopagnosic brain: toward a non-hierarchical view of face perception

    Frontiers in Human Neuroscience 4:225.

    https://doi.org/10.3389/fnhum.2010.00225

    • PubMed
    • Google Scholar
86. 1. Rossion B
    2. Prieto EA
    3. Boremanse A
    4. Kuefner D
    5. Van Belle G

    (2012) A steady-state visual evoked potential approach to individual face perception: effect of Inversion, contrast-reversal and temporal dynamics

    NeuroImage 63:1585–1600.

    https://doi.org/10.1016/j.neuroimage.2012.08.033

    • PubMed
    • Google Scholar
87. 1. Rossion B
    2. Torfs K
    3. Jacques C
    4. Liu-Shuang J

    (2015) Fast periodic presentation of natural images reveals a robust face-selective electrophysiological response in the human brain

    Journal of Vision 15:18.

    https://doi.org/10.1167/15.1.18

    • PubMed
    • Google Scholar
88. Book

    1. Rossion B
    2. Jacques C

    (2011) The N170: understanding the time-course of face perception in the human brain

    In: Luck S. J, Kappenman E. S, editors. The Oxford Handbook of Event-Related Potential Components. Oxford University Press. pp. 115–141.

    https://doi.org/10.1093/oxfordhb/9780195374148.013.0064

    • Google Scholar
89. 1. Rotshtein P
    2. Henson RN
    3. Treves A
    4. Driver J
    5. Dolan RJ

    (2005) Morphing marilyn into Maggie dissociates physical and identity face representations in the brain

    Nature Neuroscience 8:107–113.

    https://doi.org/10.1038/nn1370

    • PubMed
    • Google Scholar
90. 1. Russell R
    2. Biederman I
    3. Nederhouser M
    4. Sinha P

    (2007) The utility of surface reflectance for the recognition of upright and inverted faces

    Vision Research 47:157–165.

    https://doi.org/10.1016/j.visres.2006.11.002

    • PubMed
    • Google Scholar
91. 1. Sadeh B
    2. Podlipsky I
    3. Zhdanov A
    4. Yovel G

    (2010) Event-related potential and functional MRI measures of face-selectivity are highly correlated: a simultaneous ERP-fMRI investigation

    Human Brain Mapping 31:1490–1501.

    https://doi.org/10.1002/hbm.20952

    • PubMed
    • Google Scholar
92. 1. Schiltz C

    (2010) Holistic perception of individual faces in the right middle fusiform gyrus as evidenced by the composite face illusion

    Journal of Vision 10:1–16.

    https://doi.org/10.1167/10.2.25

    • Google Scholar
93. 1. Schwartz L
    2. Yovel G

    (2016) The roles of perceptual and conceptual information in face recognition

    Journal of Experimental Psychology: General 145:1493–1511.

    https://doi.org/10.1037/xge0000220

    • Google Scholar
94. Book

    1. Sekihara K
    2. Nagarajan SS

    (2008) Adaptive Spatial Filters for Electromagnetic Brain Imaging

    Berlin, Heidelberg: Springer.

    https://doi.org/10.1007/978-3-540-79370-0

    • Google Scholar
95. 1. Sinha P
    2. Balas B
    3. Ostrovsky Y
    4. Russell R

    (2006) Face recognition by humans: nineteen results all computer vision researchers should know about

    Proceedings of the IEEE 94:1948–1962.

    https://doi.org/10.1109/JPROC.2006.884093

    • Google Scholar
96. 1. Srinivasan R
    2. Golomb JD
    3. Martinez AM

    (2016) A neural basis of facial action recognition in humans

    Journal of Neuroscience 36:4434–4442.

    https://doi.org/10.1523/JNEUROSCI.1704-15.2016

    • PubMed
    • Google Scholar
97. 1. Steeves JK
    2. Culham JC
    3. Duchaine BC
    4. Pratesi CC
    5. Valyear KF
    6. Schindler I
    7. Humphrey GK
    8. Milner AD
    9. Goodale MA

    (2006) The fusiform face area is not sufficient for face recognition: evidence from a patient with dense prosopagnosia and no occipital face area

    Neuropsychologia 44:594–609.

    https://doi.org/10.1016/j.neuropsychologia.2005.06.013

    • PubMed
    • Google Scholar
98. 1. Sugiura M
    2. Mano Y
    3. Sasaki A
    4. Sadato N

    (2011) Beyond the memory mechanism: person-selective and nonselective processes in recognition of personally familiar faces

    Journal of Cognitive Neuroscience 23:699–715.

    https://doi.org/10.1162/jocn.2010.21469

    • PubMed
    • Google Scholar
99. 1. Tsao DY
    2. Freiwald WA
    3. Knutsen TA
    4. Mandeville JB
    5. Tootell RB

    (2003) Faces and objects in macaque cerebral cortex

    Nature Neuroscience 6:989–995.

    https://doi.org/10.1038/nn1111

    • PubMed
    • Google Scholar
100. 1. Tsao DY
     2. Freiwald WA
     3. Tootell RB
     4. Livingstone MS

     (2006) A cortical region consisting entirely of face-selective cells

     Science 311:670–674.

     https://doi.org/10.1126/science.1119983

     • PubMed
     • Google Scholar
101. 1. Tsao DY
     2. Moeller S
     3. Freiwald WA

     (2008) Comparing face patch systems in macaques and humans

     PNAS 105:19514–19519.

     https://doi.org/10.1073/pnas.0809662105

     • PubMed
     • Google Scholar
102. 1. Van Veen BD
     2. van Drongelen W
     3. Yuchtman M
     4. Suzuki A

     (1997) Localization of brain electrical activity via linearly constrained minimum variance spatial filtering

     IEEE Transactions on Biomedical Engineering 44:867–880.

     https://doi.org/10.1109/10.623056

     • PubMed
     • Google Scholar
103. 1. Wang X
     2. Chen Y
     3. Ding M

     (2008) Estimating Granger causality after stimulus onset: a cautionary note

     NeuroImage 41:767–776.

     https://doi.org/10.1016/j.neuroimage.2008.03.025

     • PubMed
     • Google Scholar
104. 1. Weibert K
     2. Flack TR
     3. Young AW
     4. Andrews TJ

     (2018) Patterns of neural response in face regions are predicted by low-level image properties

     Cortex 103:199–210.

     https://doi.org/10.1016/j.cortex.2018.03.009

     • PubMed
     • Google Scholar
105. 1. Weiner KS
     2. Barnett MA
     3. Lorenz S
     4. Caspers J
     5. Stigliani A
     6. Amunts K
     7. Zilles K
     8. Fischl B
     9. Grill-Spector K

     (2017) The cytoarchitecture of Domain-specific regions in human High-level visual cortex

     Cerebral Cortex 27:146–161.

     https://doi.org/10.1093/cercor/bhw361

     • PubMed
     • Google Scholar
106. 1. Willenbockel V
     2. Sadr J
     3. Fiset D
     4. Horne GO
     5. Gosselin F
     6. Tanaka JW

     (2010) Controlling low-level image properties: the SHINE toolbox

     Behavior Research Methods 42:671–684.

     https://doi.org/10.3758/BRM.42.3.671

     • PubMed
     • Google Scholar
107. 1. Yip AW
     2. Sinha P

     (2002) Contribution of color to face recognition

     Perception 31:995–1003.

     https://doi.org/10.1068/p3376

     • PubMed
     • Google Scholar

Author details

1. Xiaoxu Fan

   1. State Key Laboratory of Brain and Cognitive Science, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China
   2. University of Chinese Academy of Sciences, Beijing, China

   Contribution

   Conceptualization, Data curation, Formal analysis, Investigation, Visualization,Methodology, Writing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-8115-8621

2. Fan Wang

   1. State Key Laboratory of Brain and Cognitive Science, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China
   2. University of Chinese Academy of Sciences, Beijing, China

   Contribution

   Methodology, Data acquisition, Data analysis, Funding acquisition

   Competing interests

   No competing interests declared

3. Hanyu Shao

   State Key Laboratory of Brain and Cognitive Science, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China

   Contribution

   Methodology, Data acquisition

   Competing interests

   No competing interests declared

4. Peng Zhang

   1. State Key Laboratory of Brain and Cognitive Science, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China
   2. University of Chinese Academy of Sciences, Beijing, China

   Contribution

   Methodology, Data analysis, Funding acquisition

   Competing interests

   No competing interests declared

5. Sheng He

   1. State Key Laboratory of Brain and Cognitive Science, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China
   2. University of Chinese Academy of Sciences, Beijing, China
   3. Department of Psychology, University of Minnesota, Minneapolis, United States

   Contribution

   Conceptualization, Investigation, Methodology, Writing, Funding acquisition, Project administration

   For correspondence

   sheng@umn.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-5547-923X

• 12,291

  views

• 575

  downloads

• 23

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.