Siddhartha Bhattacharya
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Siddhartha Bhattacharya has a total of 12 co-authors. Here are the top co-authors by number of shared publications:
- Bosoon Park
United States Department of Agriculture, Agricultural Research Service, U.S. National Poultry Research Center, Athens, GA 30605, USA
2 shared publications - Chaaran Arunachalam
Department of Electrical and Computer Engineering Michigan State University East Lansing MI USA
2 shared publications - Jiajia Li
Department of Electrical and Computer Engineering Michigan State University East Lansing MI USA
2 shared publications - Jiyoon Yi
Department of Biosystems and Agricultural Engineering, Michigan State University, East Lansing, MI, United States
2 shared publications - Kaixiang Zhang
Department of Mechanical Engineering, Michigan State University, East Lansing, MI, USA
2 shared publications - Keyi Zhu
Department of Mechanical Engineering Michigan State University East Lansing MI USA
2 shared publications
Latest Publications
- Open Access
Rapid Salmonella Serovar Classification Using AI-Enabled Hyperspectral Microscopy with Enhanced Data Preprocessing and Multimodal Fusion
Salmonella serovar identification typically requires multiple enrichment steps using selective media, consuming considerable time and resources. This study presents a rapid, culture-independent method leveraging artificial intelligence (AI) to classify Salmonella serovars from rich hyperspectral microscopy data. Five serovars (Enteritidis, Infantis, Kentucky, Johannesburg, 4,[5],12:i:-) were analyzed from samples prepared using only sterilized de-ionized water. Hyperspectral data cubes were collected to generate single-cell spectra and RGB composite images representing the full microscopy field. Data analysis involved two parallel branches followed by multimodal fusion. The spectral branch compared manual feature selection with data-driven feature extraction via principal component analysis (PCA), followed by classification using conventional machine learning models (i.e., k-nearest neighbors, support vector machine, random forest, and multilayer perceptron). The image branch employed a convolutional neural network (CNN) to extract spatial features directly from images without predefined morphological descriptors. Using PCA-derived spectral features, the highest performing machine learning model achieved 81.1% accuracy, outperforming manual feature selection. CNN-based classification using image features alone yielded lower accuracy (57.3%) in this serovar-level discrimination. In contrast, a multimodal fusion model combining spectral and image features improved accuracy to 82.4% on the unseen test set while reducing overfitting on the train set. This study demonstrates that AI-enabled hyperspectral microscopy with multimodal fusion can streamline Salmonella serovar identification workflows.
Open AccessRapid Salmonella Serovar Classification Using AI-Enabled Hyperspectral Microscopy with Enhanced Data Preprocessing and Multimodal Fusion
Salmonella serovar identification typically requires multiple enrichment steps using selective media, consuming considerable time and resources. This study presents a rapid, culture-independent method leveraging artificial intelligence (AI) to classify Salmonella serovars from rich hyperspectral microscopy data. Five serovars (Enteritidis, Infantis, Kentucky, Johannesburg, 4,[5],12:i:-) were analyzed from samples prepared using only sterilized de-ionized water. Hyperspectral data cubes were collected to generate single-cell spectra and RGB composite images representing the full microscopy field. Data analysis involved two parallel branches followed by multimodal fusion. The spectral branch compared manual feature selection with data-driven feature extraction via principal component analysis (PCA), followed by classification using conventional machine learning models (i.e., k-nearest neighbors, support vector machine, random forest, and multilayer perceptron). The image branch employed a convolutional neural network (CNN) to extract spatial features directly from images without predefined morphological descriptors. Using PCA-derived spectral features, the highest performing machine learning model achieved 81.1% accuracy, outperforming manual feature selection. CNN-based classification using image features alone yielded lower accuracy (57.3%) in this serovar-level discrimination. In contrast, a multimodal fusion model combining spectral and image features improved accuracy to 82.4% on the unseen test set while reducing overfitting on the train set. This study demonstrates that AI-enabled hyperspectral microscopy with multimodal fusion can streamline Salmonella serovar identification workflows.
- Open Access
- 21 July 2025
- Published by MDPI AG in Preprints.org
Rapid Salmonella Serovar Classification Using AI-Enabled Hyperspectral Microscopy with Enhanced Data Preprocessing and Multimodal Fusion
Salmonella serovar identification typically requires multiple enrichment steps using selective media, consuming considerable time and resources. This study presents a rapid, culture-independent method leveraging artificial intelligence (AI) to directly classify Salmonella serovars from high-dimensional hyperspectral microscopy data. Five serovars (Enteritidis, Infantis, Kentucky, Johannesburg, 4,[5],12:i:-) were analyzed from overnight cultures prepared on glass slides using only sterilized deionized water. Per serovar, 100 hyperspectral data cubes were collected across six biological replicates (n=500 total), yielding single-cell spectra from segmented cells and RGB composite images representing the full microscopy field. Data analysis involved two parallel branches followed by multimodal fusion. The spectral branch compared two data preprocessing approaches: manual feature selection and data-driven feature extraction via principal component analysis (PCA). Single-cell spectra were classified using k-nearest neighbors, support vector machine, random forest, and multilayer perceptron (MLP), evaluated by accuracy, precision, and recall. The image branch employed a convolutional neural network (CNN) to extract spatial features directly from microscopy images without relying on explicit morphological descriptors. PCA-derived spectral features outperformed manual selection, with MLP achieving 81.1% accuracy. Multimodal MLP-CNN fusion further improved accuracy (82.4%) while reducing overfitting. This study demonstrates that AI-enabled hyperspectral microscopy with multimodal fusion can streamline Salmonella serovar identification workflows.
- 21 July 2025
- Published by MDPI AG in Preprints.org
Open AccessRapid Salmonella Serovar Classification Using AI-Enabled Hyperspectral Microscopy with Enhanced Data Preprocessing and Multimodal Fusion
Salmonella serovar identification typically requires multiple enrichment steps using selective media, consuming considerable time and resources. This study presents a rapid, culture-independent method leveraging artificial intelligence (AI) to directly classify Salmonella serovars from high-dimensional hyperspectral microscopy data. Five serovars (Enteritidis, Infantis, Kentucky, Johannesburg, 4,[5],12:i:-) were analyzed from overnight cultures prepared on glass slides using only sterilized deionized water. Per serovar, 100 hyperspectral data cubes were collected across six biological replicates (n=500 total), yielding single-cell spectra from segmented cells and RGB composite images representing the full microscopy field. Data analysis involved two parallel branches followed by multimodal fusion. The spectral branch compared two data preprocessing approaches: manual feature selection and data-driven feature extraction via principal component analysis (PCA). Single-cell spectra were classified using k-nearest neighbors, support vector machine, random forest, and multilayer perceptron (MLP), evaluated by accuracy, precision, and recall. The image branch employed a convolutional neural network (CNN) to extract spatial features directly from microscopy images without relying on explicit morphological descriptors. PCA-derived spectral features outperformed manual selection, with MLP achieving 81.1% accuracy. Multimodal MLP-CNN fusion further improved accuracy (82.4%) while reducing overfitting. This study demonstrates that AI-enabled hyperspectral microscopy with multimodal fusion can streamline Salmonella serovar identification workflows.
- Open Access
Advancement and Field Evaluation of a Dual-arm Apple Harvesting Robot
Apples are among the most widely consumed fruits worldwide. Currently, apple harvesting fully relies on manual labor, which is costly, drudging, and hazardous to workers. Hence, robotic harvesting has attracted increasing attention in recent years. However, existing systems still fall short in terms of performance, effectiveness, and reliability for complex orchard environments. In this work, we present the development and evaluation of a dual-arm harvesting robot. The system integrates a ToF camera, two 4DOF robotic arms, a centralized vacuum system, and a post-harvest handling module. During harvesting, suction force is dynamically assigned to either arm via the vacuum system, enabling efficient apple detachment while reducing power consumption and noise. Compared to our previous design, we incorporated a platform movement mechanism that enables both in-out and up-down adjustments, enhancing the robot's dexterity and adaptability to varying canopy structures. On the algorithmic side, we developed a robust apple localization pipeline that combines a foundation-model-based detector, segmentation, and clustering-based depth estimation, which improves performance in orchards. Additionally, pressure sensors were integrated into the system, and a novel dual-arm coordination strategy was introduced to respond to harvest failures based on sensor feedback, further improving picking efficiency. Field demos were conducted in two commercial orchards in MI, USA, with different canopy structures. The system achieved success rates of 0.807 and 0.797, with an average picking cycle time of 5.97s. The proposed strategy reduced harvest time by 28% compared to a single-arm baseline. The dual-arm harvesting robot enhances the reliability and efficiency of apple picking. With further advancements, the system holds strong potential for autonomous operation and commercialization for the apple industry.
Open AccessAdvancement and Field Evaluation of a Dual-arm Apple Harvesting Robot
Apples are among the most widely consumed fruits worldwide. Currently, apple harvesting fully relies on manual labor, which is costly, drudging, and hazardous to workers. Hence, robotic harvesting has attracted increasing attention in recent years. However, existing systems still fall short in terms of performance, effectiveness, and reliability for complex orchard environments. In this work, we present the development and evaluation of a dual-arm harvesting robot. The system integrates a ToF camera, two 4DOF robotic arms, a centralized vacuum system, and a post-harvest handling module. During harvesting, suction force is dynamically assigned to either arm via the vacuum system, enabling efficient apple detachment while reducing power consumption and noise. Compared to our previous design, we incorporated a platform movement mechanism that enables both in-out and up-down adjustments, enhancing the robot's dexterity and adaptability to varying canopy structures. On the algorithmic side, we developed a robust apple localization pipeline that combines a foundation-model-based detector, segmentation, and clustering-based depth estimation, which improves performance in orchards. Additionally, pressure sensors were integrated into the system, and a novel dual-arm coordination strategy was introduced to respond to harvest failures based on sensor feedback, further improving picking efficiency. Field demos were conducted in two commercial orchards in MI, USA, with different canopy structures. The system achieved success rates of 0.807 and 0.797, with an average picking cycle time of 5.97s. The proposed strategy reduced harvest time by 28% compared to a single-arm baseline. The dual-arm harvesting robot enhances the reliability and efficiency of apple picking. With further advancements, the system holds strong potential for autonomous operation and commercialization for the apple industry.
- Open Access
Foundation Model-Based Apple Ripeness and Size Estimation for Selective Harvesting
Harvesting is a critical task in the tree fruit industry, demanding extensive manual labor and substantial costs, and exposing workers to potential hazards. Recent advances in automated harvesting offer a promising solution by enabling efficient, cost-effective, and ergonomic fruit picking within tight harvesting windows. However, existing harvesting technologies often indiscriminately harvest all visible and accessible fruits, including those that are unripe or undersized. This study introduces a novel foundation model-based framework for efficient apple ripeness and size estimation. Specifically, we curated two public RGBD-based Fuji apple image datasets, integrating expanded annotations for ripeness ("Ripe" vs. "Unripe") based on fruit color and image capture dates. The resulting comprehensive dataset, Fuji-Ripeness-Size Dataset, includes 4,027 images and 16,257 annotated apples with ripeness and size labels. Using Grounding-DINO, a language-model-based object detector, we achieved robust apple detection and ripeness classification, outperforming other state-of-the-art models. Additionally, we developed and evaluated six size estimation algorithms, selecting the one with the lowest error and variation for optimal performance. The Fuji-Ripeness-Size Dataset and the apple detection and size estimation algorithms are made publicly available, which provides valuable benchmarks for future studies in automated and selective harvesting.
Open AccessFoundation Model-Based Apple Ripeness and Size Estimation for Selective Harvesting
Harvesting is a critical task in the tree fruit industry, demanding extensive manual labor and substantial costs, and exposing workers to potential hazards. Recent advances in automated harvesting offer a promising solution by enabling efficient, cost-effective, and ergonomic fruit picking within tight harvesting windows. However, existing harvesting technologies often indiscriminately harvest all visible and accessible fruits, including those that are unripe or undersized. This study introduces a novel foundation model-based framework for efficient apple ripeness and size estimation. Specifically, we curated two public RGBD-based Fuji apple image datasets, integrating expanded annotations for ripeness ("Ripe" vs. "Unripe") based on fruit color and image capture dates. The resulting comprehensive dataset, Fuji-Ripeness-Size Dataset, includes 4,027 images and 16,257 annotated apples with ripeness and size labels. Using Grounding-DINO, a language-model-based object detector, we achieved robust apple detection and ripeness classification, outperforming other state-of-the-art models. Additionally, we developed and evaluated six size estimation algorithms, selecting the one with the lowest error and variation for optimal performance. The Fuji-Ripeness-Size Dataset and the apple detection and size estimation algorithms are made publicly available, which provides valuable benchmarks for future studies in automated and selective harvesting.
- Open Access
Impact of Network Topology on Byzantine Resilience in Decentralized Federated Learning
Federated learning (FL) enables a collaborative environment for training machine learning models without sharing training data between users. This is typically achieved by aggregating model gradients on a central server. Decentralized federated learning is a rising paradigm that enables users to collaboratively train machine learning models in a peer-to-peer manner, without the need for a central aggregation server. However, before applying decentralized FL in real-world use training environments, nodes that deviate from the FL process (Byzantine nodes) must be considered when selecting an aggregation function. Recent research has focused on Byzantine-robust aggregation for client-server or fully connected networks, but has not yet evaluated such aggregation schemes for complex topologies possible with decentralized FL. Thus, the need for empirical evidence of Byzantine robustness in differing network topologies is evident. This work investigates the effects of state-of-the-art Byzantine-robust aggregation methods in complex, large-scale network structures. We find that state-of-the-art Byzantine robust aggregation strategies are not resilient within large non-fully connected networks. As such, our findings point the field towards the development of topology-aware aggregation schemes, especially necessary within the context of large scale real-world deployment.
Open AccessImpact of Network Topology on Byzantine Resilience in Decentralized Federated Learning
Federated learning (FL) enables a collaborative environment for training machine learning models without sharing training data between users. This is typically achieved by aggregating model gradients on a central server. Decentralized federated learning is a rising paradigm that enables users to collaboratively train machine learning models in a peer-to-peer manner, without the need for a central aggregation server. However, before applying decentralized FL in real-world use training environments, nodes that deviate from the FL process (Byzantine nodes) must be considered when selecting an aggregation function. Recent research has focused on Byzantine-robust aggregation for client-server or fully connected networks, but has not yet evaluated such aggregation schemes for complex topologies possible with decentralized FL. Thus, the need for empirical evidence of Byzantine robustness in differing network topologies is evident. This work investigates the effects of state-of-the-art Byzantine-robust aggregation methods in complex, large-scale network structures. We find that state-of-the-art Byzantine robust aggregation strategies are not resilient within large non-fully connected networks. As such, our findings point the field towards the development of topology-aware aggregation schemes, especially necessary within the context of large scale real-world deployment.