Kuntal Ghosh | Indian Statistical Institute, Calcutta (original) (raw)

Papers by Kuntal Ghosh

Research paper thumbnail of State-of-The-Art Fuzzy Active Contour Models for Image Segmentation

arXiv (Cornell University), Aug 1, 2020

Research paper thumbnail of A DNN based Brain Computational Model of Flicker Fusion

Research Square (Research Square), Jun 15, 2023

Research paper thumbnail of Improving Node Classification Accuracy of GNN through Input and Output Intervention

ACM Transactions on Knowledge Discovery From Data, Jul 22, 2023

Research paper thumbnail of An ensemble machine learning model based on multiple filtering and supervised attribute clustering algorithm for classifying cancer samples

PeerJ Computer Science

BackgroundMachine learning is one kind of machine intelligence technique that learns from data an... more BackgroundMachine learning is one kind of machine intelligence technique that learns from data and detects inherent patterns from large, complex datasets. Due to this capability, machine learning techniques are widely used in medical applications, especially where large-scale genomic and proteomic data are used. Cancer classification based on bio-molecular profiling data is a very important topic for medical applications since it improves the diagnostic accuracy of cancer and enables a successful culmination of cancer treatments. Hence, machine learning techniques are widely used in cancer detection and prognosis.MethodsIn this article, a new ensemble machine learning classification model named Multiple Filtering and Supervised Attribute Clustering algorithm based Ensemble Classification model (MFSAC-EC) is proposed which can handle class imbalance problem and high dimensionality of microarray datasets. This model first generates a number of bootstrapped datasets from the original t...

Research paper thumbnail of Constant community identification in million-scale networks

Social Network Analysis and Mining, Jun 30, 2022

Research paper thumbnail of Pedestrian Death and Injury in Road Accidents in Kolkata in Last Decade: A Major Public Health Challenge

Research paper thumbnail of An in-silica computation of alpha oscillations from apparently unrelated psychophysics data

The ten hertz alpha oscillations are the strongest rhythmic encodings from human brain. They aris... more The ten hertz alpha oscillations are the strongest rhythmic encodings from human brain. They arise from occipital lobes and are connected with visual perception and visual cognition. A flicker stimulus can also evoke oscillations in human EEG that are of the same frequency as the flicker stimulus. The oscillations are evoked even when there is no conscious perception of the flicker stimulus. The oscillations so evoked show resonances at 10, 20, 40 and 80 hertz. When a subject is not able to perceive the flicker and the source appears as a steady source of light, flicker fusion is said to have occurred. Psychophysics experiments on flicker fusion involves presenting a human subject with a flicker stimulus, and the subject classifying the stimulus as either flickering or fused. The physical parameters associated with the stimulus that determines whether the stimulus is classified as flickering or fused by the subject, are varied in psychophysics experiments. Deep neural networks, on t...

Research paper thumbnail of Understanding the role of starch sheath layer in graviception of Alternanthera philoxeroides: a biophysical and microscopical study

Journal of Plant Research

Research paper thumbnail of Modified Lomax model: a heavy-tailed distribution for fitting large-scale real-world complex networks

Social Network Analysis and Mining, 2021

Real-world networks are generally claimed to be scale-free, meaning that the degree distributions... more Real-world networks are generally claimed to be scale-free, meaning that the degree distributions follow the classical power-law, at least asymptotically. Yet, closer observation shows that the classical power-law distribution is often inadequate to meet the data characteristics due to the existence of a clearly identifiable non-linearity in the entire degree distribution in the log-log scale. The present paper proposes a new variant of the popular heavy-tailed Lomax distribution which we named as the Modified Lomax (MLM) distribution that can efficiently capture the crucial aspect of heavy-tailed behavior of the entire degree distribution of real-world complex networks. The proposed MLM model, derived from a hierarchical family of Lomax distributions, can efficiently fit the entire degree distribution of real-world networks without removing lower degree nodes as opposed to the classical power-law based fitting. The MLM distribution belongs to the maximum domain of attraction of the Frechet distribution and is right tail equivalent to Pareto distribution. Various statistical properties including characteristics of the maximum likelihood estimates and asymptotic distributions have also been derived for the proposed MLM model. Finally, the effectiveness of the proposed MLM model is demonstrated through rigorous experiments over fifty real-world complex networks from diverse applied domains.

Research paper thumbnail of A microscopic study on scattering in tissue section of Alternanthera philoxeroides under polarized light

Journal of Biosciences

Like any other biological tissue, plant tissue also exhibits optical properties like refraction, ... more Like any other biological tissue, plant tissue also exhibits optical properties like refraction, transmission, absorption, coloration, scattering and so on. Several studies have been conducted using different parts of plants such as leaves, seedlings, roots, stems and so on, and their optical properties have been analyzed to study plant physiology, influence of environmental cues on plant metabolism, light propagation through plant parts and the like. Thus, it is essential to study in detail the optical properties of several plant parts to determine their structural relationship. In this backdrop, an experimental study was conducted to observe and analyze the optical properties of node and inter-nodal tissue cross-sections of the plant Alternanthera philoxeroides under a polarizing microscope constructed and standardized in the laboratory. The observed optical properties of the microscopic tissue sections have been then studied to determine a significant structural relationship between nodal and inter-nodal tissue arrangement patterns as a whole. Tissue sections that have undergone a sort of biological perturbation like loss of water (dried in air for 15 min) have also been studied to study the change in the pattern of tissue optical property when compared with that of normal plant-tissue cross-sections under a polarizing microscope. This type of biological perturbation was chosen for the study because water plays an important role in maintenance of the normal physiological processes in plants and most other forms of life.

Research paper thumbnail of Linkage Between Select Anthropometric Measures and Pulmonary Function Indicator: A Study on Bengalee Male Automobile Workers

Design Science and Innovation, 2021

Assessment of respiratory functions is of considerable importance in both the physiological and d... more Assessment of respiratory functions is of considerable importance in both the physiological and diagnostic purposes, and also in occupational health in select high risk occupations. The peak expiratory flow rate (PEFR), an effective measure of effort dependent air flow, is one of the important indicators of the lung functions. It can easily and efficiently evaluate any kind of airflow obstructions and several broncho-constrictions. In this backdrop, a study has been undertaken on 39 male individuals (24–35 years) engaged in automobile works to assess their pulmonary function status and to find out and/or validate the linkage, if any, existing between select anthropometric measures and the lung function indicator. Stature (cm), body weight (kg), arm span (cm), sitting height (cm), trunk length (cm), and chest circumference (cm) were measured, and PEFR was also measured. Stature, arm span, trunk length and sitting height were significantly (P < 0.05) correlated with the indicator of the lung function-PEFR.

Research paper thumbnail of Individualized Helmet: An Approach for Reducing Road Traffic Injury Casualties

Design Science and Innovation, 2021

In India, every year, more than 4 lakhs people lose their lives in different accidents as per Gov... more In India, every year, more than 4 lakhs people lose their lives in different accidents as per Government of India data that translates to more than one person getting killed every minute in the country. Out of this, Road Traffic Injury (RTI) is the most serious preventable but large and growing public health burden especially for low and middle income countries including India. 141,526 persons were killed and 477,731 injured in road traffic accidents in India in 2014 (NCRB 2015); obviously, it includes a large number of two-wheeler riders. Motorcyclists are a group of vulnerable road users, representing 23% of the global RTI burden. It has been observed that among motorcyclists, injury to the head and neck is often the main cause of death and disability. Though use of helmets can reduce substantially the number and severity of injuries and deaths, there is common tendency to not to wear it, and the non-use or improper wearing during use of helmets is associated with injuries and disabilities in case of accidents taking place leads to higher treatment costs in the event of a crash. Studies have identified poor/ improper design of helmets as one of the reason of non-usage; therefore, addressing the underlying cause is the need of the hour. In this backdrop, a study has been undertaken to 'personalize' or tailor make helmets on the basis of anthropometric data for developing regression models to predict the appropriate helmet for the user in order to facilitate reduction of non-usage of helmets due to improper fit. Data from North and Eastern India of young adults of age range 20-30 years have been used to develop the model, and the latter has been validated subsequently. Further studies with similar focus are required to design standard 'personalized' helmets that may have the potential to reduce at least some proportion of the RTI.

Research paper thumbnail of A multi-scale Gaussian filtering model to explain the White’s illusion from the viewpoint of lightness assimilation for widely varying grating width and patch length

The variation between the actual and perceived lightness of a stimulus has strong dependency on i... more The variation between the actual and perceived lightness of a stimulus has strong dependency on its background, a phenomena commonly known as lightness induction in the literature of visual neuroscience and psychology. For instance, a gray patch may perceptually appear to be darker in a background while it looks brighter when the background is reversed. In the literature it is further reported that such variation can take place in two possible ways. In case of stimulus like the Simultaneous Brightness Contrast (SBC), the apparent lightness changes in the direction opposite to that of the background lightness, a phenomenon often referred to as lightness contrast, while in the others like pincushion or checkerboard illusion it occurs opposite to that, and known as lightness assimilation. The White’s illusion is a typical one which according to many, does not completely conform to any of these two processes. This paper presents the result of quantification of the perceptual strength of...

Research paper thumbnail of Experimental Observations of the Visibility Threshold of Illusory effects in Hermann Grid, Sinusoidal and Square Gratings and their Possible Implications

Experimental studies have been conducted on the Sinusoidal Grating, Square Grating and Hermann Gr... more Experimental studies have been conducted on the Sinusoidal Grating, Square Grating and Hermann Grid Illusions. Volunteers were presented with a range of input stimuli and asked to report the perceptibility of the illusory effect. We tried to find the thresholds for the parameters within which the illusion is visible. Some noteworthy observations were made from these experiments. We have observed that the length scale of the illusory effect changes with the typical length scale of the input stimulus, almost in proportion, wherefrom one can draw the conclusion that most spatial filters, possessing typical length scales of their own, would not respond to stimuli much smaller or much larger than that typical length scale. We have also found that even tiny changes to the pattern of the Hermann grid illusion is sufficient to wipe out the illusion completely, i.e. the illusion is very sensitive to tiny changes in the input stimulus. These observations indicate that simple linear models such as the DOG (Difference of Gaussians) model or even the highly successful ODOG (Oriented Difference of Gaussians) model of brightness perception are actually inadequate for explaining such effects. The present work thus bears implication for future efforts towards modeling the human visual system, by pointing out certain features that such models must exhibit.

Research paper thumbnail of Individualized Helmet: An Approach for Reducing Road Traffic Injury Casualties

Design Science and Innovation, 2021

In India, every year, more than 4 lakhs people lose their lives in different accidents as per Gov... more In India, every year, more than 4 lakhs people lose their lives in different accidents as per Government of India data that translates to more than one person getting killed every minute in the country. Out of this, Road Traffic Injury (RTI) is the most serious preventable but large and growing public health burden especially for low and middle income countries including India. 141,526 persons were killed and 477,731 injured in road traffic accidents in India in 2014 (NCRB 2015); obviously, it includes a large number of two-wheeler riders. Motorcyclists are a group of vulnerable road users, representing 23% of the global RTI burden. It has been observed that among motorcyclists, injury to the head and neck is often the main cause of death and disability. Though use of helmets can reduce substantially the number and severity of injuries and deaths, there is common tendency to not to wear it, and the non-use or improper wearing during use of helmets is associated with injuries and disabilities in case of accidents taking place leads to higher treatment costs in the event of a crash. Studies have identified poor/ improper design of helmets as one of the reason of non-usage; therefore, addressing the underlying cause is the need of the hour. In this backdrop, a study has been undertaken to 'personalize' or tailor make helmets on the basis of anthropometric data for developing regression models to predict the appropriate helmet for the user in order to facilitate reduction of non-usage of helmets due to improper fit. Data from North and Eastern India of young adults of age range 20-30 years have been used to develop the model, and the latter has been validated subsequently. Further studies with similar focus are required to design standard 'personalized' helmets that may have the potential to reduce at least some proportion of the RTI.

Research paper thumbnail of Deep neural networks for automatic grain-matrix segmentation in plane and cross-polarized sandstone photomicrographs

Applied Intelligence, 2021

Research paper thumbnail of A Study on Crossmodal Correspondence in Sensory Pathways Through Forced Choice Task and Frequency Based Correlation in Sound-Symbolism

Mining Intelligence and Knowledge Exploration, 2017

The crossmodal correspondence in sensory pathways of human can get revealed by subjecting them to... more The crossmodal correspondence in sensory pathways of human can get revealed by subjecting them to forced choice task as in sound-symbolism. Sound-symbolism is a term used for a hypothetical systematic relationship between word and meaning. A well known case in sound symbolism is the Kiki-Bouba phenomenon in which a subject labels a jagged figure as Kiki and rounded figure as Bouba when presented with both figures and words and asked to label the figure with the words. In the current experiment the words for cotton and sword were chosen from foreign languages and the subjects were asked to label the figure with that pair. Majority of subjects labeled the pointed figure with the word for sword for most of the languages. The word for sword had higher frequency components in most languages. The subjects may be associating words with higher frequency components to the jagged figure which implies possible crossmodal correspondence between visual and auditory pathways as was also indicated by neuropsychologists for natural language understanding.

Research paper thumbnail of Figure-Ground Segregation: Cues from Early Vision

Indian International Conference on Artificial Intelligence, 2007

Research paper thumbnail of Analysing the patterns of spatial contrast discontinuities in natural images for robust edge detection

Pattern Analysis and Applications

The pattern of spatial contrast discontinuities in natural images has been analysed in the presen... more The pattern of spatial contrast discontinuities in natural images has been analysed in the present work, and based on it, a new adaptive model of the bio-inspired Difference of Gaussian (DOG)-based edge detector has been designed. The distinguishing feature of the proposed filter is that the magnitude of surround suppression in receptive field of the DOG is adaptively adjusted depending on the nature of discontinuity of the edge profile. The model is based on the biological evidences indicating the possibility that human brain may be endowed with the ability to perform Fourier decomposition of visual images into its various components of spatial frequencies. It may be shown that information obtained from such a Fourier decomposition may help to measure the strength of contrast (sharpness of discontinuity) in the intensity profile across any possible edge in the natural image. In the present model, it is assumed that the magnitude of surround suppression in an excitatory-inhibitory receptive field is dependent on the sharpness of discontinuity. The suppression is strong when the edge contrast is poor, while it becomes weaker as the edge contrast is high. At a biphasic edge, the surround suppression is vanishingly small. Natural images collected from benchmark databases are used to evaluate the efficiency and robustness of the proposed model for the detection of edges. The result shows that the edge maps generated through the proposed model are at par, if not more effective as compared to the classical edge detectors like Canny. The performance of the proposed model is also compared with a number of recently proposed alternative adaptive models for edge detection.

Research paper thumbnail of Analysing the patterns of spatial contrast discontinuities in natural images for robust edge detection

The pattern of spatial contrast discontinuities in natural images has been analysed in the presen... more The pattern of spatial contrast discontinuities in natural images has been analysed in the present work, and based on it, a new adaptive model of the bio-inspired Difference of Gaussian (DOG)-based edge detector has been designed. The distinguishing feature of the proposed filter is that the magnitude of surround suppression in receptive field of the DOG is adaptively adjusted depending on the nature of discontinuity of the edge profile. The model is based on the biological evidences indicating the possibility that human brain may be endowed with the ability to perform Fourier decomposition of visual images into its various components of spatial frequencies. It may be shown that information obtained from such a Fourier decomposition may help to measure the strength of contrast (sharpness of discontinuity) in the intensity profile across any possible edge in the natural image. In the present model, it is assumed that the magnitude of surround suppression in an excitatory–inhibitory r...

Research paper thumbnail of State-of-The-Art Fuzzy Active Contour Models for Image Segmentation

arXiv (Cornell University), Aug 1, 2020

Research paper thumbnail of A DNN based Brain Computational Model of Flicker Fusion

Research Square (Research Square), Jun 15, 2023

Research paper thumbnail of Improving Node Classification Accuracy of GNN through Input and Output Intervention

ACM Transactions on Knowledge Discovery From Data, Jul 22, 2023

Research paper thumbnail of An ensemble machine learning model based on multiple filtering and supervised attribute clustering algorithm for classifying cancer samples

PeerJ Computer Science

BackgroundMachine learning is one kind of machine intelligence technique that learns from data an... more BackgroundMachine learning is one kind of machine intelligence technique that learns from data and detects inherent patterns from large, complex datasets. Due to this capability, machine learning techniques are widely used in medical applications, especially where large-scale genomic and proteomic data are used. Cancer classification based on bio-molecular profiling data is a very important topic for medical applications since it improves the diagnostic accuracy of cancer and enables a successful culmination of cancer treatments. Hence, machine learning techniques are widely used in cancer detection and prognosis.MethodsIn this article, a new ensemble machine learning classification model named Multiple Filtering and Supervised Attribute Clustering algorithm based Ensemble Classification model (MFSAC-EC) is proposed which can handle class imbalance problem and high dimensionality of microarray datasets. This model first generates a number of bootstrapped datasets from the original t...

Research paper thumbnail of Constant community identification in million-scale networks

Social Network Analysis and Mining, Jun 30, 2022

Research paper thumbnail of Pedestrian Death and Injury in Road Accidents in Kolkata in Last Decade: A Major Public Health Challenge

Research paper thumbnail of An in-silica computation of alpha oscillations from apparently unrelated psychophysics data

The ten hertz alpha oscillations are the strongest rhythmic encodings from human brain. They aris... more The ten hertz alpha oscillations are the strongest rhythmic encodings from human brain. They arise from occipital lobes and are connected with visual perception and visual cognition. A flicker stimulus can also evoke oscillations in human EEG that are of the same frequency as the flicker stimulus. The oscillations are evoked even when there is no conscious perception of the flicker stimulus. The oscillations so evoked show resonances at 10, 20, 40 and 80 hertz. When a subject is not able to perceive the flicker and the source appears as a steady source of light, flicker fusion is said to have occurred. Psychophysics experiments on flicker fusion involves presenting a human subject with a flicker stimulus, and the subject classifying the stimulus as either flickering or fused. The physical parameters associated with the stimulus that determines whether the stimulus is classified as flickering or fused by the subject, are varied in psychophysics experiments. Deep neural networks, on t...

Research paper thumbnail of Understanding the role of starch sheath layer in graviception of Alternanthera philoxeroides: a biophysical and microscopical study

Journal of Plant Research

Research paper thumbnail of Modified Lomax model: a heavy-tailed distribution for fitting large-scale real-world complex networks

Social Network Analysis and Mining, 2021

Real-world networks are generally claimed to be scale-free, meaning that the degree distributions... more Real-world networks are generally claimed to be scale-free, meaning that the degree distributions follow the classical power-law, at least asymptotically. Yet, closer observation shows that the classical power-law distribution is often inadequate to meet the data characteristics due to the existence of a clearly identifiable non-linearity in the entire degree distribution in the log-log scale. The present paper proposes a new variant of the popular heavy-tailed Lomax distribution which we named as the Modified Lomax (MLM) distribution that can efficiently capture the crucial aspect of heavy-tailed behavior of the entire degree distribution of real-world complex networks. The proposed MLM model, derived from a hierarchical family of Lomax distributions, can efficiently fit the entire degree distribution of real-world networks without removing lower degree nodes as opposed to the classical power-law based fitting. The MLM distribution belongs to the maximum domain of attraction of the Frechet distribution and is right tail equivalent to Pareto distribution. Various statistical properties including characteristics of the maximum likelihood estimates and asymptotic distributions have also been derived for the proposed MLM model. Finally, the effectiveness of the proposed MLM model is demonstrated through rigorous experiments over fifty real-world complex networks from diverse applied domains.

Research paper thumbnail of A microscopic study on scattering in tissue section of Alternanthera philoxeroides under polarized light

Journal of Biosciences

Like any other biological tissue, plant tissue also exhibits optical properties like refraction, ... more Like any other biological tissue, plant tissue also exhibits optical properties like refraction, transmission, absorption, coloration, scattering and so on. Several studies have been conducted using different parts of plants such as leaves, seedlings, roots, stems and so on, and their optical properties have been analyzed to study plant physiology, influence of environmental cues on plant metabolism, light propagation through plant parts and the like. Thus, it is essential to study in detail the optical properties of several plant parts to determine their structural relationship. In this backdrop, an experimental study was conducted to observe and analyze the optical properties of node and inter-nodal tissue cross-sections of the plant Alternanthera philoxeroides under a polarizing microscope constructed and standardized in the laboratory. The observed optical properties of the microscopic tissue sections have been then studied to determine a significant structural relationship between nodal and inter-nodal tissue arrangement patterns as a whole. Tissue sections that have undergone a sort of biological perturbation like loss of water (dried in air for 15 min) have also been studied to study the change in the pattern of tissue optical property when compared with that of normal plant-tissue cross-sections under a polarizing microscope. This type of biological perturbation was chosen for the study because water plays an important role in maintenance of the normal physiological processes in plants and most other forms of life.

Research paper thumbnail of Linkage Between Select Anthropometric Measures and Pulmonary Function Indicator: A Study on Bengalee Male Automobile Workers

Design Science and Innovation, 2021

Assessment of respiratory functions is of considerable importance in both the physiological and d... more Assessment of respiratory functions is of considerable importance in both the physiological and diagnostic purposes, and also in occupational health in select high risk occupations. The peak expiratory flow rate (PEFR), an effective measure of effort dependent air flow, is one of the important indicators of the lung functions. It can easily and efficiently evaluate any kind of airflow obstructions and several broncho-constrictions. In this backdrop, a study has been undertaken on 39 male individuals (24–35 years) engaged in automobile works to assess their pulmonary function status and to find out and/or validate the linkage, if any, existing between select anthropometric measures and the lung function indicator. Stature (cm), body weight (kg), arm span (cm), sitting height (cm), trunk length (cm), and chest circumference (cm) were measured, and PEFR was also measured. Stature, arm span, trunk length and sitting height were significantly (P < 0.05) correlated with the indicator of the lung function-PEFR.

Research paper thumbnail of Individualized Helmet: An Approach for Reducing Road Traffic Injury Casualties

Design Science and Innovation, 2021

In India, every year, more than 4 lakhs people lose their lives in different accidents as per Gov... more In India, every year, more than 4 lakhs people lose their lives in different accidents as per Government of India data that translates to more than one person getting killed every minute in the country. Out of this, Road Traffic Injury (RTI) is the most serious preventable but large and growing public health burden especially for low and middle income countries including India. 141,526 persons were killed and 477,731 injured in road traffic accidents in India in 2014 (NCRB 2015); obviously, it includes a large number of two-wheeler riders. Motorcyclists are a group of vulnerable road users, representing 23% of the global RTI burden. It has been observed that among motorcyclists, injury to the head and neck is often the main cause of death and disability. Though use of helmets can reduce substantially the number and severity of injuries and deaths, there is common tendency to not to wear it, and the non-use or improper wearing during use of helmets is associated with injuries and disabilities in case of accidents taking place leads to higher treatment costs in the event of a crash. Studies have identified poor/ improper design of helmets as one of the reason of non-usage; therefore, addressing the underlying cause is the need of the hour. In this backdrop, a study has been undertaken to 'personalize' or tailor make helmets on the basis of anthropometric data for developing regression models to predict the appropriate helmet for the user in order to facilitate reduction of non-usage of helmets due to improper fit. Data from North and Eastern India of young adults of age range 20-30 years have been used to develop the model, and the latter has been validated subsequently. Further studies with similar focus are required to design standard 'personalized' helmets that may have the potential to reduce at least some proportion of the RTI.

Research paper thumbnail of A multi-scale Gaussian filtering model to explain the White’s illusion from the viewpoint of lightness assimilation for widely varying grating width and patch length

The variation between the actual and perceived lightness of a stimulus has strong dependency on i... more The variation between the actual and perceived lightness of a stimulus has strong dependency on its background, a phenomena commonly known as lightness induction in the literature of visual neuroscience and psychology. For instance, a gray patch may perceptually appear to be darker in a background while it looks brighter when the background is reversed. In the literature it is further reported that such variation can take place in two possible ways. In case of stimulus like the Simultaneous Brightness Contrast (SBC), the apparent lightness changes in the direction opposite to that of the background lightness, a phenomenon often referred to as lightness contrast, while in the others like pincushion or checkerboard illusion it occurs opposite to that, and known as lightness assimilation. The White’s illusion is a typical one which according to many, does not completely conform to any of these two processes. This paper presents the result of quantification of the perceptual strength of...

Research paper thumbnail of Experimental Observations of the Visibility Threshold of Illusory effects in Hermann Grid, Sinusoidal and Square Gratings and their Possible Implications

Experimental studies have been conducted on the Sinusoidal Grating, Square Grating and Hermann Gr... more Experimental studies have been conducted on the Sinusoidal Grating, Square Grating and Hermann Grid Illusions. Volunteers were presented with a range of input stimuli and asked to report the perceptibility of the illusory effect. We tried to find the thresholds for the parameters within which the illusion is visible. Some noteworthy observations were made from these experiments. We have observed that the length scale of the illusory effect changes with the typical length scale of the input stimulus, almost in proportion, wherefrom one can draw the conclusion that most spatial filters, possessing typical length scales of their own, would not respond to stimuli much smaller or much larger than that typical length scale. We have also found that even tiny changes to the pattern of the Hermann grid illusion is sufficient to wipe out the illusion completely, i.e. the illusion is very sensitive to tiny changes in the input stimulus. These observations indicate that simple linear models such as the DOG (Difference of Gaussians) model or even the highly successful ODOG (Oriented Difference of Gaussians) model of brightness perception are actually inadequate for explaining such effects. The present work thus bears implication for future efforts towards modeling the human visual system, by pointing out certain features that such models must exhibit.

Research paper thumbnail of Individualized Helmet: An Approach for Reducing Road Traffic Injury Casualties

Design Science and Innovation, 2021

In India, every year, more than 4 lakhs people lose their lives in different accidents as per Gov... more In India, every year, more than 4 lakhs people lose their lives in different accidents as per Government of India data that translates to more than one person getting killed every minute in the country. Out of this, Road Traffic Injury (RTI) is the most serious preventable but large and growing public health burden especially for low and middle income countries including India. 141,526 persons were killed and 477,731 injured in road traffic accidents in India in 2014 (NCRB 2015); obviously, it includes a large number of two-wheeler riders. Motorcyclists are a group of vulnerable road users, representing 23% of the global RTI burden. It has been observed that among motorcyclists, injury to the head and neck is often the main cause of death and disability. Though use of helmets can reduce substantially the number and severity of injuries and deaths, there is common tendency to not to wear it, and the non-use or improper wearing during use of helmets is associated with injuries and disabilities in case of accidents taking place leads to higher treatment costs in the event of a crash. Studies have identified poor/ improper design of helmets as one of the reason of non-usage; therefore, addressing the underlying cause is the need of the hour. In this backdrop, a study has been undertaken to 'personalize' or tailor make helmets on the basis of anthropometric data for developing regression models to predict the appropriate helmet for the user in order to facilitate reduction of non-usage of helmets due to improper fit. Data from North and Eastern India of young adults of age range 20-30 years have been used to develop the model, and the latter has been validated subsequently. Further studies with similar focus are required to design standard 'personalized' helmets that may have the potential to reduce at least some proportion of the RTI.

Research paper thumbnail of Deep neural networks for automatic grain-matrix segmentation in plane and cross-polarized sandstone photomicrographs

Applied Intelligence, 2021

Research paper thumbnail of A Study on Crossmodal Correspondence in Sensory Pathways Through Forced Choice Task and Frequency Based Correlation in Sound-Symbolism

Mining Intelligence and Knowledge Exploration, 2017

The crossmodal correspondence in sensory pathways of human can get revealed by subjecting them to... more The crossmodal correspondence in sensory pathways of human can get revealed by subjecting them to forced choice task as in sound-symbolism. Sound-symbolism is a term used for a hypothetical systematic relationship between word and meaning. A well known case in sound symbolism is the Kiki-Bouba phenomenon in which a subject labels a jagged figure as Kiki and rounded figure as Bouba when presented with both figures and words and asked to label the figure with the words. In the current experiment the words for cotton and sword were chosen from foreign languages and the subjects were asked to label the figure with that pair. Majority of subjects labeled the pointed figure with the word for sword for most of the languages. The word for sword had higher frequency components in most languages. The subjects may be associating words with higher frequency components to the jagged figure which implies possible crossmodal correspondence between visual and auditory pathways as was also indicated by neuropsychologists for natural language understanding.

Research paper thumbnail of Figure-Ground Segregation: Cues from Early Vision

Indian International Conference on Artificial Intelligence, 2007

Research paper thumbnail of Analysing the patterns of spatial contrast discontinuities in natural images for robust edge detection

Pattern Analysis and Applications

The pattern of spatial contrast discontinuities in natural images has been analysed in the presen... more The pattern of spatial contrast discontinuities in natural images has been analysed in the present work, and based on it, a new adaptive model of the bio-inspired Difference of Gaussian (DOG)-based edge detector has been designed. The distinguishing feature of the proposed filter is that the magnitude of surround suppression in receptive field of the DOG is adaptively adjusted depending on the nature of discontinuity of the edge profile. The model is based on the biological evidences indicating the possibility that human brain may be endowed with the ability to perform Fourier decomposition of visual images into its various components of spatial frequencies. It may be shown that information obtained from such a Fourier decomposition may help to measure the strength of contrast (sharpness of discontinuity) in the intensity profile across any possible edge in the natural image. In the present model, it is assumed that the magnitude of surround suppression in an excitatory-inhibitory receptive field is dependent on the sharpness of discontinuity. The suppression is strong when the edge contrast is poor, while it becomes weaker as the edge contrast is high. At a biphasic edge, the surround suppression is vanishingly small. Natural images collected from benchmark databases are used to evaluate the efficiency and robustness of the proposed model for the detection of edges. The result shows that the edge maps generated through the proposed model are at par, if not more effective as compared to the classical edge detectors like Canny. The performance of the proposed model is also compared with a number of recently proposed alternative adaptive models for edge detection.

Research paper thumbnail of Analysing the patterns of spatial contrast discontinuities in natural images for robust edge detection

The pattern of spatial contrast discontinuities in natural images has been analysed in the presen... more The pattern of spatial contrast discontinuities in natural images has been analysed in the present work, and based on it, a new adaptive model of the bio-inspired Difference of Gaussian (DOG)-based edge detector has been designed. The distinguishing feature of the proposed filter is that the magnitude of surround suppression in receptive field of the DOG is adaptively adjusted depending on the nature of discontinuity of the edge profile. The model is based on the biological evidences indicating the possibility that human brain may be endowed with the ability to perform Fourier decomposition of visual images into its various components of spatial frequencies. It may be shown that information obtained from such a Fourier decomposition may help to measure the strength of contrast (sharpness of discontinuity) in the intensity profile across any possible edge in the natural image. In the present model, it is assumed that the magnitude of surround suppression in an excitatory–inhibitory r...