Leslie M. Collins
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Leslie M. Collins has a total of 94 co-authors. Here are the top co-authors by number of shared publications:
- Jordan Milton Malof
Department of Electrical Engineering and Computer Science, University of Missouri, Columbia, MO 65211, USA
10 shared publications - Boyla O. Mainsah
Department of Electrical and Computer Engineering, Duke University, Durham, NC, USA
7 shared publications Chandra S. Throckmorton
Department of Electrical and Computer Engineering Duke University 27708 NC USA
5 shared publications- Chandra S. Throckmorton
Elec. and Comput. Eng., Duke Univ., Durham, NC
4 shared publications - Kyle Bradbury
Department of Electrical and Computer Engineering, Duke University, Durham, NC 27708, USA
4 shared publications - Simiao Ren
Duke University
4 shared publications
Latest Publications
- Open Access
- 1 June 2026
- Published by MDPI AG in ISPRS International Journal of Geo-Information
- Vol. 15  (6) , 244
When Centroids Mislead: Quantifying the Consequences of Sub-Optimally Aggregating Gridded Raster Data to Polygons
Point data, such as population, disease incidence, and greenhouse gas emissions, are commonly aggregated to a uniform grid of raster data for storage and representation. In many remote sensing applications, polygons are instead used to describe regions of interest (e.g., countries and cities) which form the spatial basis for analysis. The values associated with these polygons are estimated by aggregating the underlying gridded raster data within the boundary of the polygon. The conventional approach to this aggregation relies on determining if the grid cell centroid lies within the polygon, which has accuracy limitations with potentially severe consequences. In this work, we quantify the consequence of sub-optimally aggregating gridded raster data to polygons, demonstrating that the use of the centroid alone is rarely the most accurate. Across real-world population, greenhouse gas emissions, and snowfall datasets, we further demonstrate that these aggregation-method differences emerge systematically across commonly used geographic boundaries, particularly for coarse-resolution raster datasets aggregated to county- and city-scale polygons. We compare centroid aggregation with proportional aggregation and bilinear interpolation using 2Ă— and 10Ă— upsampling. The centroid method is consistently sub-optimal, generally underperforming alternative aggregation methods when the polygon area is larger than the grid cell area. Worse, the centroid method may exhibit up to 100Ă— the error of the other aggregation methods when polygon area is no larger than grid cell area. Our findings suggest that centroid aggregation is often sub-optimal relative to alternative approaches, particularly in low-PGR settings.
- 1 June 2026
- Published by MDPI AG in ISPRS International Journal of Geo-Information
- Vol. 15  (6) , 244
Open AccessWhen Centroids Mislead: Quantifying the Consequences of Sub-Optimally Aggregating Gridded Raster Data to Polygons
Point data, such as population, disease incidence, and greenhouse gas emissions, are commonly aggregated to a uniform grid of raster data for storage and representation. In many remote sensing applications, polygons are instead used to describe regions of interest (e.g., countries and cities) which form the spatial basis for analysis. The values associated with these polygons are estimated by aggregating the underlying gridded raster data within the boundary of the polygon. The conventional approach to this aggregation relies on determining if the grid cell centroid lies within the polygon, which has accuracy limitations with potentially severe consequences. In this work, we quantify the consequence of sub-optimally aggregating gridded raster data to polygons, demonstrating that the use of the centroid alone is rarely the most accurate. Across real-world population, greenhouse gas emissions, and snowfall datasets, we further demonstrate that these aggregation-method differences emerge systematically across commonly used geographic boundaries, particularly for coarse-resolution raster datasets aggregated to county- and city-scale polygons. We compare centroid aggregation with proportional aggregation and bilinear interpolation using 2Ă— and 10Ă— upsampling. The centroid method is consistently sub-optimal, generally underperforming alternative aggregation methods when the polygon area is larger than the grid cell area. Worse, the centroid method may exhibit up to 100Ă— the error of the other aggregation methods when polygon area is no larger than grid cell area. Our findings suggest that centroid aggregation is often sub-optimal relative to alternative approaches, particularly in low-PGR settings.
- Open Access
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 155146-155203
A Cross-Domain Survey of Medical and Remote Sensing Image Registration Methods
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 155146-155203
Open AccessA Cross-Domain Survey of Medical and Remote Sensing Image Registration Methods
- Open Access
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 19, 1-29
Data-Efficient Deep Learning for SAR Automatic Target Recognition: A Comprehensive Survey
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 19, 1-29
Open AccessData-Efficient Deep Learning for SAR Automatic Target Recognition: A Comprehensive Survey
- 29 July 2025
- Published by Elsevier in Pattern Recognition
- Vol. 171
Are deep learning models robust to partial object occlusion in visual recognition tasks?
- 29 July 2025
- Published by Elsevier in Pattern Recognition
- Vol. 171
Are deep learning models robust to partial object occlusion in visual recognition tasks?
- Open Access
Are Deep Learning Models Robust to Partial Object Occlusion in Visual Recognition Tasks?
Image classification models, including convolutional neural networks (CNNs), perform well on a variety of classification tasks but struggle under conditions of partial occlusion, i.e., conditions in which objects are partially covered from the view of a camera. Methods to improve performance under occlusion, including data augmentation, part-based clustering, and more inherently robust architectures, including Vision Transformer (ViT) models, have, to some extent, been evaluated on their ability to classify objects under partial occlusion. However, evaluations of these methods have largely relied on images containing artificial occlusion, which are typically computer-generated and therefore inexpensive to label. Additionally, methods are rarely compared against each other, and many methods are compared against early, now outdated, deep learning models. We contribute the Image Recognition Under Occlusion (IRUO) dataset, based on the recently developed Occluded Video Instance Segmentation (OVIS) dataset (arXiv:2102.01558). IRUO utilizes real-world and artificially occluded images to test and benchmark leading methods' robustness to partial occlusion in visual recognition tasks. In addition, we contribute the design and results of a human study using images from IRUO that evaluates human classification performance at multiple levels and types of occlusion. We find that modern CNN-based models show improved recognition accuracy on occluded images compared to earlier CNN-based models, and ViT-based models are more accurate than CNN-based models on occluded images, performing only modestly worse than human accuracy. We also find that certain types of occlusion, including diffuse occlusion, where relevant objects are seen through "holes" in occluders such as fences and leaves, can greatly reduce the accuracy of deep recognition models as compared to humans, especially those with CNN backbones.
Open AccessAre Deep Learning Models Robust to Partial Object Occlusion in Visual Recognition Tasks?
Image classification models, including convolutional neural networks (CNNs), perform well on a variety of classification tasks but struggle under conditions of partial occlusion, i.e., conditions in which objects are partially covered from the view of a camera. Methods to improve performance under occlusion, including data augmentation, part-based clustering, and more inherently robust architectures, including Vision Transformer (ViT) models, have, to some extent, been evaluated on their ability to classify objects under partial occlusion. However, evaluations of these methods have largely relied on images containing artificial occlusion, which are typically computer-generated and therefore inexpensive to label. Additionally, methods are rarely compared against each other, and many methods are compared against early, now outdated, deep learning models. We contribute the Image Recognition Under Occlusion (IRUO) dataset, based on the recently developed Occluded Video Instance Segmentation (OVIS) dataset (arXiv:2102.01558). IRUO utilizes real-world and artificially occluded images to test and benchmark leading methods' robustness to partial occlusion in visual recognition tasks. In addition, we contribute the design and results of a human study using images from IRUO that evaluates human classification performance at multiple levels and types of occlusion. We find that modern CNN-based models show improved recognition accuracy on occluded images compared to earlier CNN-based models, and ViT-based models are more accurate than CNN-based models on occluded images, performing only modestly worse than human accuracy. We also find that certain types of occlusion, including diffuse occlusion, where relevant objects are seen through "holes" in occluders such as fences and leaves, can greatly reduce the accuracy of deep recognition models as compared to humans, especially those with CNN backbones.
- Open Access
Segment anything, from space?
Recently, the first foundation model developed specifically for image segmentation tasks was developed, termed the "Segment Anything Model" (SAM). SAM can segment objects in input imagery based on cheap input prompts, such as one (or more) points, a bounding box, or a mask. The authors examined the \textit{zero-shot} image segmentation accuracy of SAM on a large number of vision benchmark tasks and found that SAM usually achieved recognition accuracy similar to, or sometimes exceeding, vision models that had been trained on the target tasks. The impressive generalization of SAM for segmentation has major implications for vision researchers working on natural imagery. In this work, we examine whether SAM's performance extends to overhead imagery problems and help guide the community's response to its development. We examine SAM's performance on a set of diverse and widely studied benchmark tasks. We find that SAM does often generalize well to overhead imagery, although it fails in some cases due to the unique characteristics of overhead imagery and its common target objects. We report on these unique systematic failure cases for remote sensing imagery that may comprise useful future research for the community.
Open AccessSegment anything, from space?
Recently, the first foundation model developed specifically for image segmentation tasks was developed, termed the "Segment Anything Model" (SAM). SAM can segment objects in input imagery based on cheap input prompts, such as one (or more) points, a bounding box, or a mask. The authors examined the \textit{zero-shot} image segmentation accuracy of SAM on a large number of vision benchmark tasks and found that SAM usually achieved recognition accuracy similar to, or sometimes exceeding, vision models that had been trained on the target tasks. The impressive generalization of SAM for segmentation has major implications for vision researchers working on natural imagery. In this work, we examine whether SAM's performance extends to overhead imagery problems and help guide the community's response to its development. We examine SAM's performance on a set of diverse and widely studied benchmark tasks. We find that SAM does often generalize well to overhead imagery, although it fails in some cases due to the unique characteristics of overhead imagery and its common target objects. We report on these unique systematic failure cases for remote sensing imagery that may comprise useful future research for the community.
- Open Access
- 26 June 2023
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 37  (8) , 9874-9881
Mixture Manifold Networks: A Computationally Efficient Baseline for Inverse Modeling
We propose and show the efficacy of a new method to address generic inverse problems. Inverse modeling is the task whereby one seeks to determine the hidden parameters of a natural system that produce a given set of observed measurements. Recent work has shown impressive results using deep learning, but we note that there is a trade-off between model performance and computational time. For some applications, the computational time at inference for the best performing inverse modeling method may be overly prohibitive to its use. In seeking a faster, high-performing model, we present a new method that leverages multiple manifolds as a mixture of backward (e.g., inverse) models in a forward-backward model architecture. These multiple backwards models all share a common forward model, and their training is mitigated by generating training examples from the forward model. The proposed method thus has two innovations: 1) the multiple Manifold Mixture Network (MMN) architecture, and 2) the training procedure involving augmenting backward model training data using the forward model. We demonstrate the advantages of our method by comparing to several baselines on four benchmark inverse problems, and we furthermore provide analysis to motivate its design.
- 26 June 2023
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 37  (8) , 9874-9881
Open AccessMixture Manifold Networks: A Computationally Efficient Baseline for Inverse Modeling
We propose and show the efficacy of a new method to address generic inverse problems. Inverse modeling is the task whereby one seeks to determine the hidden parameters of a natural system that produce a given set of observed measurements. Recent work has shown impressive results using deep learning, but we note that there is a trade-off between model performance and computational time. For some applications, the computational time at inference for the best performing inverse modeling method may be overly prohibitive to its use. In seeking a faster, high-performing model, we present a new method that leverages multiple manifolds as a mixture of backward (e.g., inverse) models in a forward-backward model architecture. These multiple backwards models all share a common forward model, and their training is mitigated by generating training examples from the forward model. The proposed method thus has two innovations: 1) the multiple Manifold Mixture Network (MMN) architecture, and 2) the training procedure involving augmenting backward model training data using the forward model. We demonstrate the advantages of our method by comparing to several baselines on four benchmark inverse problems, and we furthermore provide analysis to motivate its design.
- Open Access
Mixture Manifold Networks: A Computationally Efficient Baseline for Inverse Modeling
We propose and show the efficacy of a new method to address generic inverse problems. Inverse modeling is the task whereby one seeks to determine the control parameters of a natural system that produce a given set of observed measurements. Recent work has shown impressive results using deep learning, but we note that there is a trade-off between model performance and computational time. For some applications, the computational time at inference for the best performing inverse modeling method may be overly prohibitive to its use. We present a new method that leverages multiple manifolds as a mixture of backward (e.g., inverse) models in a forward-backward model architecture. These multiple backwards models all share a common forward model, and their training is mitigated by generating training examples from the forward model. The proposed method thus has two innovations: 1) the multiple Manifold Mixture Network (MMN) architecture, and 2) the training procedure involving augmenting backward model training data using the forward model. We demonstrate the advantages of our method by comparing to several baselines on four benchmark inverse problems, and we furthermore provide analysis to motivate its design.
Open AccessMixture Manifold Networks: A Computationally Efficient Baseline for Inverse Modeling
We propose and show the efficacy of a new method to address generic inverse problems. Inverse modeling is the task whereby one seeks to determine the control parameters of a natural system that produce a given set of observed measurements. Recent work has shown impressive results using deep learning, but we note that there is a trade-off between model performance and computational time. For some applications, the computational time at inference for the best performing inverse modeling method may be overly prohibitive to its use. We present a new method that leverages multiple manifolds as a mixture of backward (e.g., inverse) models in a forward-backward model architecture. These multiple backwards models all share a common forward model, and their training is mitigated by generating training examples from the forward model. The proposed method thus has two innovations: 1) the multiple Manifold Mixture Network (MMN) architecture, and 2) the training procedure involving augmenting backward model training data using the forward model. We demonstrate the advantages of our method by comparing to several baselines on four benchmark inverse problems, and we furthermore provide analysis to motivate its design.
- Open Access
Meta-simulation for the Automated Design of Synthetic Overhead Imagery
The use of synthetic (or simulated) data for training machine learning models has grown rapidly in recent years. Synthetic data can often be generated much faster and more cheaply than its real-world counterpart. One challenge of using synthetic imagery however is scene design: e.g., the choice of content and its features and spatial arrangement. To be effective, this design must not only be realistic, but appropriate for the target domain, which (by assumption) is unlabeled. In this work, we propose an approach to automatically choose the design of synthetic imagery based upon unlabeled real-world imagery. Our approach, termed Neural-Adjoint Meta-Simulation (NAMS), builds upon the seminal recent meta-simulation approaches. In contrast to the current state-of-the-art methods, our approach can be pre-trained once offline, and then provides fast design inference for new target imagery. Using both synthetic and real-world problems, we show that NAMS infers synthetic designs that match both the in-domain and out-of-domain target imagery, and that training segmentation models with NAMS-designed imagery yields superior results compared to na\"ive randomized designs and state-of-the-art meta-simulation methods.
Open AccessMeta-simulation for the Automated Design of Synthetic Overhead Imagery
The use of synthetic (or simulated) data for training machine learning models has grown rapidly in recent years. Synthetic data can often be generated much faster and more cheaply than its real-world counterpart. One challenge of using synthetic imagery however is scene design: e.g., the choice of content and its features and spatial arrangement. To be effective, this design must not only be realistic, but appropriate for the target domain, which (by assumption) is unlabeled. In this work, we propose an approach to automatically choose the design of synthetic imagery based upon unlabeled real-world imagery. Our approach, termed Neural-Adjoint Meta-Simulation (NAMS), builds upon the seminal recent meta-simulation approaches. In contrast to the current state-of-the-art methods, our approach can be pre-trained once offline, and then provides fast design inference for new target imagery. Using both synthetic and real-world problems, we show that NAMS infers synthetic designs that match both the in-domain and out-of-domain target imagery, and that training segmentation models with NAMS-designed imagery yields superior results compared to na\"ive randomized designs and state-of-the-art meta-simulation methods.
- 23 July 2022
- Published by Taylor & Francis in Cochlear Implants International
- Vol. 23  (6) , 309-316
Parameter tuning of time-frequency masking algorithms for reverberant artifact removal within the cochlear implant stimulus
Cochlear implant recipients struggle to understand speech in reverberant environments. To restore speech perception, artifacts due to reverberant reflections can be removed from the cochlear implant stimulus by applying a matrix of gain values, a technique referred to as time-frequency masking. In this study, two common time-frequency masking strategies are implemented within cochlear implant processing, either introducing complete retention or deletion of stimulus components using a binary mask or continuous attenuation of stimulus components using a ratio mask. Parameters of each masking strategy control the level of attenuation imposed by the gain values. In this study, we perceptually tune the parameters of the masking strategy to determine a balance between speech retention and artifact removal. We measure the intelligibility of reverberant signals mitigated by each strategy with speech recognition testing in normal-hearing listeners using vocoding as a simulation of cochlear implant perception. For both masking strategies, we find parameterizations that maximize the intelligibility of the mitigated signals. At the best-performing parameterizations, binary-masked reverberant signals yield larger intelligibility improvements than ratio-masked signals. The results provide a perceptually optimized objective for the removal of reverberant artifacts from cochlear implant stimuli, facilitating improved speech recognition performance for cochlear implant recipients in reverberant environments.
- 23 July 2022
- Published by Taylor & Francis in Cochlear Implants International
- Vol. 23  (6) , 309-316
Parameter tuning of time-frequency masking algorithms for reverberant artifact removal within the cochlear implant stimulus
Cochlear implant recipients struggle to understand speech in reverberant environments. To restore speech perception, artifacts due to reverberant reflections can be removed from the cochlear implant stimulus by applying a matrix of gain values, a technique referred to as time-frequency masking. In this study, two common time-frequency masking strategies are implemented within cochlear implant processing, either introducing complete retention or deletion of stimulus components using a binary mask or continuous attenuation of stimulus components using a ratio mask. Parameters of each masking strategy control the level of attenuation imposed by the gain values. In this study, we perceptually tune the parameters of the masking strategy to determine a balance between speech retention and artifact removal. We measure the intelligibility of reverberant signals mitigated by each strategy with speech recognition testing in normal-hearing listeners using vocoding as a simulation of cochlear implant perception. For both masking strategies, we find parameterizations that maximize the intelligibility of the mitigated signals. At the best-performing parameterizations, binary-masked reverberant signals yield larger intelligibility improvements than ratio-masked signals. The results provide a perceptually optimized objective for the removal of reverberant artifacts from cochlear implant stimuli, facilitating improved speech recognition performance for cochlear implant recipients in reverberant environments.