chintan mandal | Jadavpur University, Kolkata, India (original) (raw)

Papers by chintan mandal

Research paper thumbnail of A revisit to the planar subdivision graph: free space detection in a dynamic environment with polygonal obstacles

International Journal of Information Technology, Jan 5, 2024

Research paper thumbnail of Non-intersecting curved paths using Bezier curves in the 2D-Euclidean plane for multiple autonomous robots

International Journal of Computational Vision and Robotics, 2022

Research paper thumbnail of Bengali Text Classification: A New multi-class Dataset and Performance Evaluation of Machine Learning and Deep Learning Models

This study focuses on Bengali text classification using machine learning and deep learning techni... more This study focuses on Bengali text classification using machine learning and deep learning techniques. Text classification is a fundamental task in natural language processing (NLP) that involves categorizing text documents into predefined categories or classes. While text classification has received considerable attention in the English language, there is a limited amount of research specifically addressing Bengali text classification. This gap in the literature highlights the need for exploring and developing effective techniques tailored to the Bengali language. Furthermore, although some datasets for Bengali text classification tasks have been produced, most of the datasets have a limited number of labels. In our work, we introduce a new dataset for the Bengali text classification task, which has 38 class labels. The dataset includes data from the leading Bengali newspapers. We have evaluated many state-of-the-art machine learning and deep learning classification methods on our ...

Research paper thumbnail of Prediction of behavior of a noncommunicative robot based on observations of its movement

Concurrency and Computation: Practice and Experience

An adversary robot can move around in a free space “patrolling an area” appearing to have a rando... more An adversary robot can move around in a free space “patrolling an area” appearing to have a random “behavior” if observed through a “lens.” The “behavior” of the robot is associated with its observed/internal state. Stochastic techniques or machine learning techniques, which can be used to predict the future states of an adversary robot demand high computation and thus, power. Advanced RNN techniques like long short‐term memory, which can make predictions using long‐term dependencies of the sequence data, also have extremely high computing and memory requirements. Thus, the systems involved in using these former techniques are not appropriate to be deployed in fields which require inconspicuous monitoring. In this work, we present two novel methods using the “behavior” to predict the future states of an adversary robot which can be deployed using a low computing device/small single board computer/robot in a non‐GPS environment without any internet connection. The proposed two method...

Research paper thumbnail of Non-intersecting curved paths using Bezier curves in the 2D-Euclidean plane for multiple autonomous robots

International Journal of Computational Vision and Robotics

Research paper thumbnail of A Method for Finding Multiple Large Rectangular Free Spaces in a Map with Convex and Concave Obstacles

2021 IEEE 18th India Council International Conference (INDICON), 2021

Research paper thumbnail of Maintaining a Minimum Spanning Tree for Kinetic Autonomous Robots in 2 D-Euclidean Plane

This paper presents a procedure for maintaining a 2D-Euclidean Minimum Spanning Tree for a set of... more This paper presents a procedure for maintaining a 2D-Euclidean Minimum Spanning Tree for a set of n autonomous kinetic robots having no central supervision. The proposed procedure is based on the Kinetic data structure framework and the well known fact that the edges of the minimum spanning tree for a given set of points in the 2D-Euclidean domain are contained in the edges of its Delaunay triangulation. The kinetic data structure framework has a centralized data structure along with a priority queue on which a proposed algorithm works to maintain a combinatorial geometrical structure for a set of geometric objects. In this work, we propose an approach where the computation of the geometrical structure is done through the geometrical objects, in our case, the computations being done by the robots themselves. This is unlike that of maintaining a centralized data structure in the kinetic data structure framework. The kinetic autonomous robots, each being a computing object do all comp...

Research paper thumbnail of Accurate Prediction of Dimensionless Vector and Scalar Propagation Constants of Single-Mode Graded Index Fiber in the Low V Region by a Simple Method

We prescribe simple but accurate analytical expressionfor scalar and vector propagation constants... more We prescribe simple but accurate analytical expressionfor scalar and vector propagation constants for single-mode graded index fibers in the low V region. Our formulation is based on the power series expression of the fundamental mode of graded index fiber derived by Chebyshev formalism. Choosing parabolic index fiber as a typical example of graded index fiber, we employ the prescribed formulations in order to evaluate dimensionless vector and scalar propagation constants in the low V region. The execution of our formalism requires little computations.

Research paper thumbnail of A novel and accurate method for analysis of single-mode dispersion-shifted and dispersion-flattened fiber directional coupler

Optik, 2018

Employing the simple series expression for fundamental mode based on Chebyshev technique, we pres... more Employing the simple series expression for fundamental mode based on Chebyshev technique, we present a simple but accurate analysis of coupling characteristics of identical single-mode dispersion-shifted trapezoidal and dispersion-flattened graded W and step W fibers directional couplers. Analytical expressions of the relevant parameters are formulated and the concerned calculations require simple and little computations. We show that our formulations on coupling characteristics of such type of fibers directional couplers provide excellent results.

Research paper thumbnail of Prediction of First Higher Order Modal Field for Graded Index Fiber in Presence of Kerr Nonlinearity

Journal of Optical Communications, 2018

We report evaluation of first higher order modal field for dual mode optical fiber having step an... more We report evaluation of first higher order modal field for dual mode optical fiber having step and parabolic index profiles. The study is carried out both in absence as well as in presence of Kerr nonlinearity. The analysis is based on a simple iterative method involving Chebyshev formalism. Taking some typical step- and parabolic-index fibers as examples, we show that our results agree excellently with the exact results which can be obtained by applying rigorous methods. Thus, our simple formalism stands the merit of being considered as an accurate alternative to the existing cumbersome methods. The prescribed formalism provides scope for accurate estimation of different propagation parameters associated with first higher order mode in such kinds of fibers in presence of Kerr nonlinearity. The execution of formalism being user friendly, it will be beneficial to the system engineers working in the field of optical technology.

Research paper thumbnail of Prediction of Fundamental Modal Field for Graded Index Fiber in the Presence of Kerr Nonlinearity

Journal of Optical Communications, 2018

The power series formulation for modal field of single-mode graded index fibers by Chebyshev tech... more The power series formulation for modal field of single-mode graded index fibers by Chebyshev technique has worked excellently in predicting accurately different propagation characteristics in simple fashion. Here we develop a simple iterative method involving Chebyshev formalism to predict the modal field of single-mode graded index fiber in the presence of Kerr-type nonlinearity. Taking step and parabolic index fibers as typical examples, we show that our results match excellently with the available exact results obtained vigorously. Thus, the reported technique can be considered as an accurate alternative to the existing cumbersome techniques. Accordingly, this formalism will be beneficial to the technologies for evaluation of modal noise in single-mode Kerr-type nonlinear graded index fibers.

Research paper thumbnail of Laser Diode to Single-Mode Circular Core Parabolic Index Fiber Coupling via Upside-Down Tapered Hyperbolic Microlens on the Tip of the Fiber: Prediction of Coupling Optics by ABCD Matrix Formalism

Journal of Optical Communications, 2017

We employ ABCD matrix formalism in order to investigate the coupling optics involving laser diode... more We employ ABCD matrix formalism in order to investigate the coupling optics involving laser diode to single-mode circular core parabolic index fiber excitation via upside-down tapered hyperbolic microlens on the fiber tip. Analytic expressions for coupling efficiencies both in absence and in presence of transverse and angular mismatches are formulated. The concerned investigations are made for two practical wavelengths namely 1.3 µm and 1.5 µm. The execution of the prescribed formulations involves little computation. It has been found that the wavelength 1.5 µm is more efficient in respect of coupling. It is also seen that the present coupling device at both the wavelengths shows more tolerance with respect to angular mismatch. As regards tolerance with respect to transverse mismatch, the result is poor at both the wavelengths used. Consequently, it is desirable that designers should not to exceed transverse mismatch beyond 1 μm.

Research paper thumbnail of Approximate Minimum Spanning Tree for Points Moving in a Euclidean Two-Dimensions Plane

Lecture Notes in Electrical Engineering, 2013

The Minimum Spanning Tree (MST) is an extensively studied problem and has various graph and geome... more The Minimum Spanning Tree (MST) is an extensively studied problem and has various graph and geometric based applications. For a connected and undirected graph \(G(V,E)\) with positive edge weights, a minimum spanning tree (MST) is an acyclic subgraph of \(G\) which connects all the vertices in \(G\) such that the total edge weight is minimum. The Euclidean Minimum Spanning Tree (EMST) is a special case of MST where the Euclidean distance between the vertices are taken as edge weight.

Research paper thumbnail of Online Delaunay Triangulation Using the Quad-Edge Data Structure

Communications in Computer and Information Science, 2011

Previous works involving the Online Delaunay triangulation problem required that the incoming req... more Previous works involving the Online Delaunay triangulation problem required that the incoming request lies within the triangulation or a predefined initial triangulation framework, which will contain all the incoming points. No mention is made for Online Delaunay triangulation when the request point lies outside the triangulation, which also happens to be the unbounded side of the Convex Hull of the triangulation. In this work, we give a solution for the Online Delaunay triangulation Problem for incoming request points lying in the unbounded side of the Convex Hull bounding the Delaunay triangulation as well for points lying inside the triangulation. We use the Quad-Edge data structure for implementing the Delaunay triangulation.

Research paper thumbnail of Map Merging for an Autonomous Multi-robotic System

Advances in Intelligent Systems and Computing, 2021

Map-merging technique is a well-studied and challenging domain in the field of autonomous mobile ... more Map-merging technique is a well-studied and challenging domain in the field of autonomous mobile robotics. It is not always possible to create a map only using a single robot. Thus, the most common practice is to collect maps of individual robots of the region it has been able to reveal, also known as sub-maps and merge all of them to create a single map. In this paper, we propose an algorithm for merging sub-maps acquired from different autonomous mobile robots. We also show that the merging of the maps is dependent on the degrees of freedom of the robots.

Research paper thumbnail of A revisit to the planar subdivision graph: free space detection in a dynamic environment with polygonal obstacles

International Journal of Information Technology, Jan 5, 2024

Research paper thumbnail of Non-intersecting curved paths using Bezier curves in the 2D-Euclidean plane for multiple autonomous robots

International Journal of Computational Vision and Robotics, 2022

Research paper thumbnail of Bengali Text Classification: A New multi-class Dataset and Performance Evaluation of Machine Learning and Deep Learning Models

This study focuses on Bengali text classification using machine learning and deep learning techni... more This study focuses on Bengali text classification using machine learning and deep learning techniques. Text classification is a fundamental task in natural language processing (NLP) that involves categorizing text documents into predefined categories or classes. While text classification has received considerable attention in the English language, there is a limited amount of research specifically addressing Bengali text classification. This gap in the literature highlights the need for exploring and developing effective techniques tailored to the Bengali language. Furthermore, although some datasets for Bengali text classification tasks have been produced, most of the datasets have a limited number of labels. In our work, we introduce a new dataset for the Bengali text classification task, which has 38 class labels. The dataset includes data from the leading Bengali newspapers. We have evaluated many state-of-the-art machine learning and deep learning classification methods on our ...

Research paper thumbnail of Prediction of behavior of a noncommunicative robot based on observations of its movement

Concurrency and Computation: Practice and Experience

An adversary robot can move around in a free space “patrolling an area” appearing to have a rando... more An adversary robot can move around in a free space “patrolling an area” appearing to have a random “behavior” if observed through a “lens.” The “behavior” of the robot is associated with its observed/internal state. Stochastic techniques or machine learning techniques, which can be used to predict the future states of an adversary robot demand high computation and thus, power. Advanced RNN techniques like long short‐term memory, which can make predictions using long‐term dependencies of the sequence data, also have extremely high computing and memory requirements. Thus, the systems involved in using these former techniques are not appropriate to be deployed in fields which require inconspicuous monitoring. In this work, we present two novel methods using the “behavior” to predict the future states of an adversary robot which can be deployed using a low computing device/small single board computer/robot in a non‐GPS environment without any internet connection. The proposed two method...

Research paper thumbnail of Non-intersecting curved paths using Bezier curves in the 2D-Euclidean plane for multiple autonomous robots

International Journal of Computational Vision and Robotics

Research paper thumbnail of A Method for Finding Multiple Large Rectangular Free Spaces in a Map with Convex and Concave Obstacles

2021 IEEE 18th India Council International Conference (INDICON), 2021

Research paper thumbnail of Maintaining a Minimum Spanning Tree for Kinetic Autonomous Robots in 2 D-Euclidean Plane

This paper presents a procedure for maintaining a 2D-Euclidean Minimum Spanning Tree for a set of... more This paper presents a procedure for maintaining a 2D-Euclidean Minimum Spanning Tree for a set of n autonomous kinetic robots having no central supervision. The proposed procedure is based on the Kinetic data structure framework and the well known fact that the edges of the minimum spanning tree for a given set of points in the 2D-Euclidean domain are contained in the edges of its Delaunay triangulation. The kinetic data structure framework has a centralized data structure along with a priority queue on which a proposed algorithm works to maintain a combinatorial geometrical structure for a set of geometric objects. In this work, we propose an approach where the computation of the geometrical structure is done through the geometrical objects, in our case, the computations being done by the robots themselves. This is unlike that of maintaining a centralized data structure in the kinetic data structure framework. The kinetic autonomous robots, each being a computing object do all comp...

Research paper thumbnail of Accurate Prediction of Dimensionless Vector and Scalar Propagation Constants of Single-Mode Graded Index Fiber in the Low V Region by a Simple Method

We prescribe simple but accurate analytical expressionfor scalar and vector propagation constants... more We prescribe simple but accurate analytical expressionfor scalar and vector propagation constants for single-mode graded index fibers in the low V region. Our formulation is based on the power series expression of the fundamental mode of graded index fiber derived by Chebyshev formalism. Choosing parabolic index fiber as a typical example of graded index fiber, we employ the prescribed formulations in order to evaluate dimensionless vector and scalar propagation constants in the low V region. The execution of our formalism requires little computations.

Research paper thumbnail of A novel and accurate method for analysis of single-mode dispersion-shifted and dispersion-flattened fiber directional coupler

Optik, 2018

Employing the simple series expression for fundamental mode based on Chebyshev technique, we pres... more Employing the simple series expression for fundamental mode based on Chebyshev technique, we present a simple but accurate analysis of coupling characteristics of identical single-mode dispersion-shifted trapezoidal and dispersion-flattened graded W and step W fibers directional couplers. Analytical expressions of the relevant parameters are formulated and the concerned calculations require simple and little computations. We show that our formulations on coupling characteristics of such type of fibers directional couplers provide excellent results.

Research paper thumbnail of Prediction of First Higher Order Modal Field for Graded Index Fiber in Presence of Kerr Nonlinearity

Journal of Optical Communications, 2018

We report evaluation of first higher order modal field for dual mode optical fiber having step an... more We report evaluation of first higher order modal field for dual mode optical fiber having step and parabolic index profiles. The study is carried out both in absence as well as in presence of Kerr nonlinearity. The analysis is based on a simple iterative method involving Chebyshev formalism. Taking some typical step- and parabolic-index fibers as examples, we show that our results agree excellently with the exact results which can be obtained by applying rigorous methods. Thus, our simple formalism stands the merit of being considered as an accurate alternative to the existing cumbersome methods. The prescribed formalism provides scope for accurate estimation of different propagation parameters associated with first higher order mode in such kinds of fibers in presence of Kerr nonlinearity. The execution of formalism being user friendly, it will be beneficial to the system engineers working in the field of optical technology.

Research paper thumbnail of Prediction of Fundamental Modal Field for Graded Index Fiber in the Presence of Kerr Nonlinearity

Journal of Optical Communications, 2018

The power series formulation for modal field of single-mode graded index fibers by Chebyshev tech... more The power series formulation for modal field of single-mode graded index fibers by Chebyshev technique has worked excellently in predicting accurately different propagation characteristics in simple fashion. Here we develop a simple iterative method involving Chebyshev formalism to predict the modal field of single-mode graded index fiber in the presence of Kerr-type nonlinearity. Taking step and parabolic index fibers as typical examples, we show that our results match excellently with the available exact results obtained vigorously. Thus, the reported technique can be considered as an accurate alternative to the existing cumbersome techniques. Accordingly, this formalism will be beneficial to the technologies for evaluation of modal noise in single-mode Kerr-type nonlinear graded index fibers.

Research paper thumbnail of Laser Diode to Single-Mode Circular Core Parabolic Index Fiber Coupling via Upside-Down Tapered Hyperbolic Microlens on the Tip of the Fiber: Prediction of Coupling Optics by ABCD Matrix Formalism

Journal of Optical Communications, 2017

We employ ABCD matrix formalism in order to investigate the coupling optics involving laser diode... more We employ ABCD matrix formalism in order to investigate the coupling optics involving laser diode to single-mode circular core parabolic index fiber excitation via upside-down tapered hyperbolic microlens on the fiber tip. Analytic expressions for coupling efficiencies both in absence and in presence of transverse and angular mismatches are formulated. The concerned investigations are made for two practical wavelengths namely 1.3 µm and 1.5 µm. The execution of the prescribed formulations involves little computation. It has been found that the wavelength 1.5 µm is more efficient in respect of coupling. It is also seen that the present coupling device at both the wavelengths shows more tolerance with respect to angular mismatch. As regards tolerance with respect to transverse mismatch, the result is poor at both the wavelengths used. Consequently, it is desirable that designers should not to exceed transverse mismatch beyond 1 μm.

Research paper thumbnail of Approximate Minimum Spanning Tree for Points Moving in a Euclidean Two-Dimensions Plane

Lecture Notes in Electrical Engineering, 2013

The Minimum Spanning Tree (MST) is an extensively studied problem and has various graph and geome... more The Minimum Spanning Tree (MST) is an extensively studied problem and has various graph and geometric based applications. For a connected and undirected graph \(G(V,E)\) with positive edge weights, a minimum spanning tree (MST) is an acyclic subgraph of \(G\) which connects all the vertices in \(G\) such that the total edge weight is minimum. The Euclidean Minimum Spanning Tree (EMST) is a special case of MST where the Euclidean distance between the vertices are taken as edge weight.

Research paper thumbnail of Online Delaunay Triangulation Using the Quad-Edge Data Structure

Communications in Computer and Information Science, 2011

Previous works involving the Online Delaunay triangulation problem required that the incoming req... more Previous works involving the Online Delaunay triangulation problem required that the incoming request lies within the triangulation or a predefined initial triangulation framework, which will contain all the incoming points. No mention is made for Online Delaunay triangulation when the request point lies outside the triangulation, which also happens to be the unbounded side of the Convex Hull of the triangulation. In this work, we give a solution for the Online Delaunay triangulation Problem for incoming request points lying in the unbounded side of the Convex Hull bounding the Delaunay triangulation as well for points lying inside the triangulation. We use the Quad-Edge data structure for implementing the Delaunay triangulation.

Research paper thumbnail of Map Merging for an Autonomous Multi-robotic System

Advances in Intelligent Systems and Computing, 2021

Map-merging technique is a well-studied and challenging domain in the field of autonomous mobile ... more Map-merging technique is a well-studied and challenging domain in the field of autonomous mobile robotics. It is not always possible to create a map only using a single robot. Thus, the most common practice is to collect maps of individual robots of the region it has been able to reveal, also known as sub-maps and merge all of them to create a single map. In this paper, we propose an algorithm for merging sub-maps acquired from different autonomous mobile robots. We also show that the merging of the maps is dependent on the degrees of freedom of the robots.