How does information from low and high spatial frequencies interact during scene categorization?

Kauffmann, Louise; Roux-Sibilon, Alexia; Beffara, Brice; Mermillod, Martial; Guyader, Nathalie; Peyrin, Carole

doi:10.1080/13506285.2017.1347590

Cited by 17 publications

(20 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Scene perception is understood as involving a coarse-to-fine processing sequence using both low spatial frequency cues (rapidly processed and allowing for parsing of global structure) and high frequency information (which is relayed more slowly to high-level areas; Kauffmann et al., 2014 ). The present study links the rapid processing of low frequency cues to the formation of categorical representations, supporting previous reports of coarse visual analysis as rapid and crucial to gist perception ( Kauffmann et al., 2017 ; Peyrin et al., 2010 ; Schyns and Oliva, 1994 ). On the other hand, high spatial frequency representations of scenes do not generalize to unfiltered stimuli, suggesting that they may encode low-level differences rather than a categorical response.…”

Section: Discussionsupporting

confidence: 90%

Spatial frequency supports the emergence of categorical representations in visual cortex during natural scene perception

2018

View full text Add to dashboard Cite

In navigating our environment, we rapidly process and extract meaning from visual cues. However, the relationship between visual features and categorical representations in natural scene perception is still not well understood. Here, we used natural scene stimuli from different categories and filtered at different spatial frequencies to address this question in a passive viewing paradigm. Using representational similarity analysis (RSA) and cross-decoding of magnetoencephalography (MEG) data, we show that categorical representations emerge in human visual cortex at ∼180 ms and are linked to spatial frequency processing. Furthermore, dorsal and ventral stream areas reveal temporally and spatially overlapping representations of low and high-level layer activations extracted from a feedforward neural network. Our results suggest that neural patterns from extrastriate visual cortex switch from low-level to categorical representations within 200 ms, highlighting the rapid cascade of processing stages essential in human visual perception.

show abstract

Section: Discussionsupporting

confidence: 90%

Spatial frequency supports the emergence of categorical representations in visual cortex during natural scene perception

2018

View full text Add to dashboard Cite

show abstract

“…Although the base architecture splits at the third stage of a CNN (e.g., ResNet), it does not mean that the two separate branches should never communicate with each other. As [11], [40] show that there exists connections and communications between LH and RH in the human brain, our LB and GB should also communicate with each other so that the features from two branches can complement and enrich each other. It will not get worse if we selectively leverage useful knowledge learned by the other branch to help the task of the current branch.…”

Section: B Gated Intercommunication Unitsmentioning

confidence: 99%

“…Although these deep learning models have advanced one step towards hemispheric specialization, they are still primitive in interhemispheric communications. As early as 1954, Tomasch and Joseph had already proposed that there exists connections and communications between LH and RH [40]; they also stated that this connective structure in the human brain is corpus callosum, which is the largest white matter structure in the human brain and consists of 200-300 million axonal projections [28]; Kauffmann et al also have indicated that LH and RH influence each other [11]. The corpus callosum plays an essential role in brain signal processing, e.g., it allows us to identify the objects we see by connecting the visual cortex with the language centers of the brain, and it also transfers tactile information between the brain hemispheres to enable us to locate touch.…”

Section: Introductionmentioning

confidence: 99%

Dual Intercommunication Network: Enabling Interhemispheric Communications in Hemisphere-Inspired ANNs

2020

IEEE Access

View full text Add to dashboard Cite

The human brain has been a main source of inspiration for designing deep learning models. Recently, inspired by the specialized functions of two cerebral hemispheres in processing low and high spatial frequency information, some dual-path neural networks with global and local branches have been proposed to deal with both coarse-and fine-grained visual tasks simultaneously. However, in existing works, the interhemispheric communication mechanism, which is responded by the corpus callosum, the largest white matter structure in the human brain that connecting the left and right cerebral hemispheres, is still not fully explored and exploited. This paper aims to explore how the corpus callosum can inspire us to enable transfer and integration of information between global and local branches in hemisphere-inspired artificial neural networks, such that one branch can leverage the other's learned knowledge and benefit each other. To this end, we propose a gated intercommunication unit to selectively transfer useful knowledge between the two branches via attention mechanisms to alleviate the negative transfer. Experiments on sb-MNIST and two pedestrian attribute datasets show that the proposed method outperforms the compared ones in most cases. INDEX TERMS Attribute recognition, computer vision, deep learning, multi-task learning.

show abstract

“…Stimuli were 160 black-and-white photographs of scenes (256-level grayscales, 256 × 256 pixels) from the Labelme database (Oliva & Torralba, 2001), previously used in works of Kauffmann et al (2017) and Kauffmann, Bourgin, et al (2015). Scenes belonged to two distinct semantic categories: 80 man-made scenes (buildings, streets, highways) and 80 natural scenes (beach, open countryside, forests).…”

Section: Stimulimentioning

confidence: 99%

Semantic and Physical Properties of Peripheral Vision Are Used for Scene Categorization in Central Vision

Peyrin

Roux-Sibilon

Trouilloud

et al. 2021

Journal of Cognitive Neuroscience

Self Cite

View full text Add to dashboard Cite

Theories of visual recognition postulate that our ability to understand our visual environment at a glance is based on the extraction of the gist of the visual scene, a first global and rudimentary visual representation. Gist perception would be based on the rapid analysis of low spatial frequencies in the visual signal and would allow a coarse categorization of the scene. We aimed to study whether the low spatial resolution information available in peripheral vision could modulate the processing of visual information presented in central vision. We combined behavioral measures (Experiments 1 and 2) and fMRI measures (Experiment 2). Participants categorized a scene presented in central vision (artificial vs. natural categories) while ignoring another scene, either semantically congruent or incongruent, presented in peripheral vision. The two scenes could either share the same physical properties (similar amplitude spectrum and spatial configuration) or not. Categorization of the central scene was impaired by a semantically incongruent peripheral scene, in particular when the two scenes were physically similar. This semantic interference effect was associated with increased activation of the inferior frontal gyrus. When the two scenes were semantically congruent, the dissimilarity of their physical properties impaired the categorization of the central scene. This effect was associated with increased activation in occipito-temporal areas. In line with the hypothesis of predictive mechanisms involved in visual recognition, results suggest that semantic and physical properties of the information coming from peripheral vision would be automatically used to generate predictions that guide the processing of signal in central vision.

show abstract

How does information from low and high spatial frequencies interact during scene categorization?

Cited by 17 publications

References 47 publications

Spatial frequency supports the emergence of categorical representations in visual cortex during natural scene perception

Spatial frequency supports the emergence of categorical representations in visual cortex during natural scene perception

Dual Intercommunication Network: Enabling Interhemispheric Communications in Hemisphere-Inspired ANNs

Semantic and Physical Properties of Peripheral Vision Are Used for Scene Categorization in Central Vision

Contact Info

Product

Resources

About