Coarse-to-fine information integration in human vision

Petras, Kirsten; Oever, Sanne ten; Jacobs, Christianne; Goffaux, Valérie

doi:10.1016/j.neuroimage.2018.10.086

Cited by 56 publications

(44 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We saw the dominance of the same features, with the most pronounced decoding values for the 'Face' category. This is consistent with previous finding showing stronger coding of face category information in the human brain compared to other categories of objects (Kaneshiro et al, 2015;Karimi-Rouzbahani et al, 2017a;Petras et al, 2019). In order to quantify the advantage for the face category, we performed a signed-rank statistical testing to quantitatively compare the decoding results across categories.…”

Section: Resultssupporting

confidence: 88%

Informative Neural Codes to Separate Object Categories

Shahmohammadi

Vahab

Karimi-Rouzbahani

2020

Preprint

View full text Add to dashboard Cite

In order to develop object recognition algorithms, which can approach human-level recognition performance, researchers have been studying how the human brain performs recognition in the past five decades. This has already in-spired AI-based object recognition algorithms, such as convolutional neural networks, which are among the most successful object recognition platforms today and can approach human performance in specific tasks. However, it is not yet clearly known how recorded brain activations convey information about object category processing. One main obstacle has been the lack of large feature sets, to evaluate the information contents of multiple aspects of neural activations. Here, we compared the information contents of a large set of 25 features, extracted from time series of electroencephalography (EEG) recorded from human participants doing an object recognition task. We could characterize the most informative aspects of brain activations about object categories. Among the evaluated features, event-related potential (ERP) components of N1 and P2a were among the most informative features with the highest information in the Theta frequency bands. Upon limiting the analysis time window, we observed more information for features detecting temporally informative patterns in the signals. The results of this study can constrain previous theories about how the brain codes object category information.

show abstract

Section: Resultssupporting

confidence: 88%

Informative Neural Codes to Separate Object Categories

Shahmohammadi

Vahab

Karimi-Rouzbahani

2020

Preprint

View full text Add to dashboard Cite

show abstract

“…and pasted onto a uniform grey background. To ensure that the image background shared the face's spectral properties, the whole image was phasescrambled in Fourier space, the original face was pasted back onto the now-scrambled background, and then the entire image was phase-scrambled again [37]. This procedure of pasting the face-related pixels back onto the background and phase-scrambling the resultant image was repeated 500 times ( Fig 1A).…”

Section: Stimulimentioning

confidence: 99%

Contrast versus identity encoding in the face image follow distinct orientation selectivity profiles

Jacobs¹,

Petras

Moors

et al. 2020

PLoS ONE

Self Cite

View full text Add to dashboard Cite

Orientation selectivity is a fundamental property of primary visual encoding. High-level processing stages also show some form of orientation dependence, with face identification preferentially relying on horizontally-oriented information. How high-level orientation tuning emerges from primary orientation biases is unclear. In the same group of participants, we derived the orientation selectivity profile at primary and high-level visual processing stages using a contrast detection and an identity matching task. To capture the orientation selectivity profile, we calculated the difference in performance between all tested orientations (0, 45, 90, and 135˚) for each task and for upright and inverted faces, separately. Primary orientation selectivity was characterized by higher sensitivity to oblique as compared to cardinal orientations. The orientation profile of face identification showed superior horizontal sensitivity to face identity. In each task, performance with upright and inverted faces projected onto qualitatively similar a priori models of orientation selectivity. Yet the fact that the orientation selectivity profiles of contrast detection in upright and inverted faces correlated significantly while such correlation was absent for identification indicates a progressive dissociation of orientation selectivity profiles from primary to high-level stages of orientation encoding. Bayesian analyses further indicate a lack of correlation between the orientation selectivity profiles in the contrast detection and face identification tasks, for upright and inverted faces. From these findings, we conclude that orientation selectivity shows distinct profiles at primary and high-level stages of face processing and that a transformation must occur from general cardinal attenuation when processing basic properties of the face image to horizontal tuning when encoding more complex properties such as identity.

show abstract

“…Activation of orbitofrontal cortex (OFC) is elicited by stimuli containing LSF information, 50ms prior to areas in the temporal cortex, but not by HSF-only images, indicating that OFC has an important role in top-down facilitation of image recognition (Bar et al, 2006). LSF modulates HSF processing in broadband images, as measured by EEG in human participants (Petras et al, 2019). When LSF is informative of the image content, HSF contributes to a lesser extent, indicating that coarse information does indeed guide the processing of fine detail.…”

Section: Introductionmentioning

confidence: 99%

Training for object recognition with increasing spatial frequency: A comparison of deep learning with human vision

Avberšek

Zeman

Beeck

2021

Preprint

View full text Add to dashboard Cite

The ontogenetic development of human vision, and the real-time neural processing of visual input, both exhibit a striking similarity – a sensitivity towards spatial frequencies that progress in a coarse-to-fine manner. During early human development, sensitivity for higher spatial frequencies increases with age. In adulthood, when humans receive new visual input, low spatial frequencies are typically processed first before subsequently guiding the processing of higher spatial frequencies. We investigated to what extent this coarse-to-fine progression might impact visual representations in artificial vision and compared this to adult human representations. We simulated the coarse-to-fine progression of image processing in deep convolutional neural networks (CNNs) by gradually increasing spatial frequency information during training. We compared CNN performance, after standard and coarse-to-fine training, with a wide range of datasets from behavioural and neuroimaging experiments. In contrast to humans, CNNs that are trained using the standard protocol are very insensitive to low spatial frequency information, showing very poor performance in being able to classify such object images. By training CNNs using our coarse-to-fine method, we improved the classification accuracy of CNNs from 0% to 32% on low-pass filtered images taken from the ImageNet dataset. When comparing differently trained networks on images containing full spatial frequency information, we saw no representational differences. Overall, this integration of computational, neural, and behavioural findings shows the relevance of the exposure to and processing of input with a variation in spatial frequency content for some aspects of high-level object representations.

show abstract

Coarse-to-fine information integration in human vision

Cited by 56 publications

References 51 publications

Informative Neural Codes to Separate Object Categories

Informative Neural Codes to Separate Object Categories

Contrast versus identity encoding in the face image follow distinct orientation selectivity profiles

Training for object recognition with increasing spatial frequency: A comparison of deep learning with human vision

Contact Info

Product

Resources

About