-
OXSeg: Multidimensional attention UNet-based lip segmentation using semi-supervised lip contours
Authors:
Hanie Moghaddasi,
Christina Chambers,
Sarah N. Mattson,
Jeffrey R. Wozniak,
Claire D. Coles,
Raja Mukherjee,
Michael Suttie
Abstract:
Lip segmentation plays a crucial role in various domains, such as lip synchronization, lipreading, and diagnostics. However, the effectiveness of supervised lip segmentation is constrained by the availability of lip contour in the training phase. A further challenge with lip segmentation is its reliance on image quality , lighting, and skin tone, leading to inaccuracies in the detected boundaries.…
▽ More
Lip segmentation plays a crucial role in various domains, such as lip synchronization, lipreading, and diagnostics. However, the effectiveness of supervised lip segmentation is constrained by the availability of lip contour in the training phase. A further challenge with lip segmentation is its reliance on image quality , lighting, and skin tone, leading to inaccuracies in the detected boundaries. To address these challenges, we propose a sequential lip segmentation method that integrates attention UNet and multidimensional input. We unravel the micro-patterns in facial images using local binary patterns to build multidimensional inputs. Subsequently, the multidimensional inputs are fed into sequential attention UNets, where the lip contour is reconstructed. We introduce a mask generation method that uses a few anatomical landmarks and estimates the complete lip contour to improve segmentation accuracy. This mask has been utilized in the training phase for lip segmentation. To evaluate the proposed method, we use facial images to segment the upper lips and subsequently assess lip-related facial anomalies in subjects with fetal alcohol syndrome (FAS). Using the proposed lip segmentation method, we achieved a mean dice score of 84.75%, and a mean pixel accuracy of 99.77% in upper lip segmentation. To further evaluate the method, we implemented classifiers to identify those with FAS. Using a generative adversarial network (GAN), we reached an accuracy of 98.55% in identifying FAS in one of the study populations. This method could be used to improve lip segmentation accuracy, especially around Cupid's bow, and shed light on distinct lip-related characteristics of FAS.
△ Less
Submitted 8 May, 2025;
originally announced May 2025.
-
Propeller Motion of a Devil-Stick using Normal Forcing
Authors:
Aakash Khandelwal,
Ranjan Mukherjee
Abstract:
The problem of realizing rotary propeller motion of a devil-stick in the vertical plane using forces purely normal to the stick is considered. This problem represents a nonprehensile manipulation task of an underactuated system. In contrast with previous approaches, the devil-stick is manipulated by controlling the normal force and its point of application. Virtual holonomic constraints are used t…
▽ More
The problem of realizing rotary propeller motion of a devil-stick in the vertical plane using forces purely normal to the stick is considered. This problem represents a nonprehensile manipulation task of an underactuated system. In contrast with previous approaches, the devil-stick is manipulated by controlling the normal force and its point of application. Virtual holonomic constraints are used to design the trajectory of the center-of-mass of the devil-stick in terms of its orientation angle, and conditions for stable propeller motion are derived. Intermittent large-amplitude forces are used to asymptotically stabilize a desired propeller motion. Simulations demonstrate the efficacy of the approach in realizing stable propeller motion without loss of contact between the actuator and devil-stick.
△ Less
Submitted 29 January, 2025;
originally announced January 2025.
-
Modeling Human Strategy for Flattening Wrinkled Cloth Using Neural Networks
Authors:
Nilay Kant,
Ashrut Aryal,
Rajiv Ranganathan,
Ranjan Mukherjee,
Charles Owen
Abstract:
This paper explores a novel approach to model strategies for flattening wrinkled cloth learning from humans. A human participant study was conducted where the participants were presented with various wrinkle types and tasked with flattening the cloth using the fewest actions possible. A camera and Aruco marker were used to capture images of the cloth and finger movements, respectively. The human s…
▽ More
This paper explores a novel approach to model strategies for flattening wrinkled cloth learning from humans. A human participant study was conducted where the participants were presented with various wrinkle types and tasked with flattening the cloth using the fewest actions possible. A camera and Aruco marker were used to capture images of the cloth and finger movements, respectively. The human strategies for flattening the cloth were modeled using a supervised regression neural network, where the cloth images served as input and the human actions as output. Before training the neural network, a series of image processing techniques were applied, followed by Principal Component Analysis (PCA) to extract relevant features from each image and reduce the input dimensionality. This reduction decreased the model's complexity and computational cost. The actions predicted by the neural network closely matched the actual human actions on an independent data set, demonstrating the effectiveness of neural networks in modeling human actions for flattening wrinkled cloth.
△ Less
Submitted 19 August, 2024;
originally announced September 2024.
-
Stabilization of Energy-Conserving Gaits for Point-Foot Planar Bipeds
Authors:
Aakash Khandelwal,
Nilay Kant,
Ranjan Mukherjee
Abstract:
The problem of designing and stabilizing impact-free, energy-conserving gaits is considered for underactuated, point-foot planar bipeds. Virtual holonomic constraints are used to design energy-conserving gaits. A desired gait corresponds to a periodic hybrid orbit and is stabilized using the Impulse Controlled Poincaré Map approach. Numerical simulations for the case of a five-link biped demonstra…
▽ More
The problem of designing and stabilizing impact-free, energy-conserving gaits is considered for underactuated, point-foot planar bipeds. Virtual holonomic constraints are used to design energy-conserving gaits. A desired gait corresponds to a periodic hybrid orbit and is stabilized using the Impulse Controlled Poincaré Map approach. Numerical simulations for the case of a five-link biped demonstrate convergence to a desired gait from arbitrary initial conditions.
△ Less
Submitted 26 October, 2022;
originally announced October 2022.
-
Nonprehensile Manipulation of a Stick Using Impulsive Forces
Authors:
Aakash Khandelwal,
Nilay Kant,
Ranjan Mukherjee
Abstract:
The problem of nonprehensile manipulation of a stick in three-dimensional space using intermittent impulsive forces is considered. The objective is to juggle the stick between a sequence of configurations that are rotationally symmetric about the vertical axis. The dynamics of the stick is described by five generalized coordinates and three control inputs. Between two consecutive configurations wh…
▽ More
The problem of nonprehensile manipulation of a stick in three-dimensional space using intermittent impulsive forces is considered. The objective is to juggle the stick between a sequence of configurations that are rotationally symmetric about the vertical axis. The dynamics of the stick is described by five generalized coordinates and three control inputs. Between two consecutive configurations where impulsive inputs are applied, the dynamics is conveniently represented by a Poincaré map in the reference frame of the juggler. Stabilization of the orbit associated with a desired juggling motion is accomplished by stabilizing a fixed point on the Poincaré map. The Impulse Controlled Poincaré Map approach is used to stabilize the orbit, and numerical simulations are used to demonstrate convergence to the desired juggling motion from an arbitrary initial configuration. In the limiting case, where consecutive rotationally symmetric configurations are chosen arbitrarily close, it is shown that the dynamics reduces to that of steady precession of the stick on a hoop.
△ Less
Submitted 6 July, 2022; v1 submitted 11 February, 2022;
originally announced February 2022.
-
Deep Learning based Framework for Automatic Diagnosis of Glaucoma based on analysis of Focal Notching in the Optic Nerve Head
Authors:
Sneha Dasgupta,
Rishav Mukherjee,
Kaushik Dutta,
Anindya Sen
Abstract:
Automatic evaluation of the retinal fundus image is emerging as one of the most important tools for early detection and treatment of progressive eye diseases like Glaucoma. Glaucoma results to a progressive degeneration of vision and is characterized by the deformation of the shape of optic cup and the degeneration of the blood vessels resulting in the formation of a notch along the neuroretinal r…
▽ More
Automatic evaluation of the retinal fundus image is emerging as one of the most important tools for early detection and treatment of progressive eye diseases like Glaucoma. Glaucoma results to a progressive degeneration of vision and is characterized by the deformation of the shape of optic cup and the degeneration of the blood vessels resulting in the formation of a notch along the neuroretinal rim. In this paper, we propose a deep learning-based pipeline for automatic segmentation of optic disc (OD) and optic cup (OC) regions from Digital Fundus Images (DFIs), thereby extracting distinct features necessary for prediction of Glaucoma. This methodology has utilized focal notch analysis of neuroretinal rim along with cup-to-disc ratio values as classifying parameters to enhance the accuracy of Computer-aided design (CAD) systems in analyzing glaucoma. Support Vector-based Machine Learning algorithm is used for classification, which classifies DFIs as Glaucomatous or Normal based on the extracted features. The proposed pipeline was evaluated on the freely available DRISHTI-GS dataset with a resultant accuracy of 93.33% for detecting Glaucoma from DFIs.
△ Less
Submitted 10 December, 2021;
originally announced December 2021.
-
Feedback-Induced Flutter Instability of a Flexible Beam in Fluid Flow
Authors:
Sanders Aspelund,
Ranjan Mukherjee,
Aren Hellum
Abstract:
A pinned-free beam in axial fluid flow, subjected to feedback-based actuation at the pinned end, is investigated. The actuation may be a moment or a prescribed angle and it is proportional to the state (curvature, slope, or displacement) of the beam at some point along its length. All equations and boundary condition terms are non-dimensionalized and the stability of the system is studied over a r…
▽ More
A pinned-free beam in axial fluid flow, subjected to feedback-based actuation at the pinned end, is investigated. The actuation may be a moment or a prescribed angle and it is proportional to the state (curvature, slope, or displacement) of the beam at some point along its length. All equations and boundary condition terms are non-dimensionalized and the stability of the system is studied over a range of external flow velocity and sensing location. For each combination of flow velocity and sensing location, the critical gain (positive or negative) for the onset of flutter is determined. This process, which is repeated for each combination of actuation and sensing modes, reveals that the closed-loop system exhibits a rich set of stability transitions, each associated with a traveling waveform in the flexible beam at the onset of flutter. With the intent of exploring the use of flexible fluttering beams for underwater propulsion, the efficiency of these waveforms is computed using slender-body theory. Additional insights into the efficiency of the waveforms are obtained through considerations of the smoothness of the traveling waveforms.
△ Less
Submitted 4 February, 2023; v1 submitted 27 October, 2021;
originally announced October 2021.
-
Orbital Stabilization of Underactuated Systems using Virtual Holonomic Constraints and Impulse Controlled Poincare Maps
Authors:
Nilay Kant,
Ranjan Mukherjee
Abstract:
The problem of orbital stabilization of underactuated mechanical systems with one passive degree-of-freedom (DOF) is revisited. Virtual holonomic constraints are enforced using a continuous controller; this results in a dense set of closed orbits on a constraint manifold. A desired orbit is selected on the manifold and a Poincare section is constructed at a fixed point on the orbit. The correspond…
▽ More
The problem of orbital stabilization of underactuated mechanical systems with one passive degree-of-freedom (DOF) is revisited. Virtual holonomic constraints are enforced using a continuous controller; this results in a dense set of closed orbits on a constraint manifold. A desired orbit is selected on the manifold and a Poincare section is constructed at a fixed point on the orbit. The corresponding Poincare map is linearized about the fixed point; this results in a discrete linear time-invariant system. To stabilize the desired orbit, impulsive inputs are applied when the system trajectory crosses the Poincare section; these inputs can be designed using standard techniques such as LQR. The Impulse Controlled Poincare Map (ICPM) based control design has lower complexity and computational cost than control designs proposed earlier. The generality of the ICPM approach is demonstrated using the 2-DOF cart-pendulum and the 3-DOF tiptoebot.
△ Less
Submitted 2 May, 2020;
originally announced May 2020.
-
Planar Symmetric Juggling of a Devil-Stick
Authors:
Nilay Kant,
Ranjan Mukherjee
Abstract:
Juggling a devil-stick can be described as a problem of non-prehensile manipulation. Assuming that the devil-stick remains confined to the vertical plane, the problem of juggling the stick between two symmetric configurations is considered. Impulsive forces are applied to the stick intermittently and the impulse of the force and its point of application are modeled as control inputs to the system.…
▽ More
Juggling a devil-stick can be described as a problem of non-prehensile manipulation. Assuming that the devil-stick remains confined to the vertical plane, the problem of juggling the stick between two symmetric configurations is considered. Impulsive forces are applied to the stick intermittently and the impulse of the force and its point of application are modeled as control inputs to the system. The dynamics of the devil-stick due to the impulsive forces and gravity is described by half-return maps between two Poincare sections; the symmetric configurations are fixed points of these sections. A coordinate transformation is used to convert the juggling problem to that of stabilization of one of the fixed points. Inclusion of the coordinate transformation in the dynamic model results in a nonlinear discrete-time system. A dead-beat design for one of the inputs simplifies the control problem and results in a linear time-invariant discrete-time system. Standard control techniques are used to show that symmetric juggling can be achieved from arbitrary initial conditions.
△ Less
Submitted 25 July, 2020; v1 submitted 1 May, 2020;
originally announced May 2020.
-
StationPlot: A New Non-stationarity Quantification Tool for Detection of Epileptic Seizures
Authors:
Sawon Pratiher,
Subhankar Chattoraj,
Rajdeep Mukherjee
Abstract:
A novel non-stationarity visualization tool known as StationPlot is developed for deciphering the chaotic behavior of a dynamical time series. A family of analytic measures enumerating geometrical aspects of the non-stationarity & degree of variability is formulated by convex hull geometry (CHG) on StationPlot. In the Euclidean space, both trend-stationary (TS) & difference-stationary (DS) perturb…
▽ More
A novel non-stationarity visualization tool known as StationPlot is developed for deciphering the chaotic behavior of a dynamical time series. A family of analytic measures enumerating geometrical aspects of the non-stationarity & degree of variability is formulated by convex hull geometry (CHG) on StationPlot. In the Euclidean space, both trend-stationary (TS) & difference-stationary (DS) perturbations are comprehended by the asymmetric structure of StationPlot's region of interest (ROI). The proposed method is experimentally validated using EEG signals, where it comprehend the relative temporal evolution of neural dynamics & its non-stationary morphology, thereby exemplifying its diagnostic competence for seizure activity (SA) detection. Experimental results & analysis-of-Variance (ANOVA) on the extracted CHG features demonstrates better classification performances as compared to the existing shallow feature based state-of-the-art & validates its efficacy as geometry-rich discriminative descriptors for signal processing applications.
△ Less
Submitted 10 November, 2018;
originally announced November 2018.
-
A Review of Micromachined Thermal Accelerometers
Authors:
Rahul Mukherjee,
Joydeep Basu,
Pradip Mandal,
Prasanta Kumar Guha
Abstract:
Thermal convection based micro-electromechanical accelerometer is a relatively new kind of acceleration sensor that does not require a solid proof mass, yielding unique benefits like high shock survival rating, low production cost, and integrability with CMOS integrated circuit technology. This article provides a comprehensive survey of the research, development, and current trends in the field of…
▽ More
Thermal convection based micro-electromechanical accelerometer is a relatively new kind of acceleration sensor that does not require a solid proof mass, yielding unique benefits like high shock survival rating, low production cost, and integrability with CMOS integrated circuit technology. This article provides a comprehensive survey of the research, development, and current trends in the field of thermal acceleration sensors, with detailed enumeration on the theory, operation, modeling, and numerical simulation of such devices. Different reported varieties and structures of thermal accelerometers have been reviewed highlighting key design, implementation, and performance aspects. Materials and technologies used for fabrication of such sensors have also been discussed. Further, the advantages and challenges for thermal accelerometers vis-à-vis other prominent accelerometer types have been presented, followed by an overview of associated signal conditioning circuitry and potential applications.
△ Less
Submitted 11 January, 2018;
originally announced January 2018.
