Pramit Saha
YOU?
Author Swipe
View article: Between Strength and Stillness: The Tale of Myotonia in an Adult Warrior: A Case Report
Between Strength and Stillness: The Tale of Myotonia in an Adult Warrior: A Case Report Open
Myotonic Dystrophy type 1 (DM1) is an autosomal dominant multisystem disorder characterized by progressive muscle weakness, myotonia, and involvement of multiple organ systems, with variable clinical presentations. We present the case of a…
View article: FedAgentBench: Towards Automating Real-world Federated Medical Image Analysis with Server-Client LLM Agents
FedAgentBench: Towards Automating Real-world Federated Medical Image Analysis with Server-Client LLM Agents Open
Federated learning (FL) allows collaborative model training across healthcare sites without sharing sensitive patient data. However, real-world FL deployment is often hindered by complex operational challenges that demand substantial human…
View article: Neural Collapse-Inspired Multi-Label Federated Learning under Label-Distribution Skew
Neural Collapse-Inspired Multi-Label Federated Learning under Label-Distribution Skew Open
Federated Learning (FL) enables collaborative model training across distributed clients while preserving data privacy, yet it remains challenging as data distributions can be highly heterogeneous. These challenges are further amplified in …
View article: Self-supervised Learning of Echocardiographic Video Representations via Online Cluster Distillation
Self-supervised Learning of Echocardiographic Video Representations via Online Cluster Distillation Open
Self-supervised learning (SSL) has achieved major advances in natural images and video understanding, but challenges remain in domains like echocardiography (heart ultrasound) due to subtle anatomical structures, complex temporal dynamics,…
View article: FedPIA – Permuting and Integrating Adapters Leveraging Wasserstein Barycenters for Finetuning Foundation Models in Multi-Modal Federated Learning
FedPIA – Permuting and Integrating Adapters Leveraging Wasserstein Barycenters for Finetuning Foundation Models in Multi-Modal Federated Learning Open
Large Vision-Language Models (VLMs), possessing millions or billions of parameters, typically require large text and image datasets for effective fine-tuning. However, collecting data from various sites, especially in healthcare, is challe…
View article: Incongruent Multimodal Federated Learning for Medical Vision and Language-based Multi-label Disease Detection
Incongruent Multimodal Federated Learning for Medical Vision and Language-based Multi-label Disease Detection Open
Federated Learning (FL) in healthcare ensures patient privacy by allowing hospitals to collaboratively train machine learning models while keeping sensitive medical data secure and localized. Most existing research in FL has concentrated o…
View article: MCAT: Visual Query-Based Localization of Standard Anatomical Clips in Fetal Ultrasound Videos Using Multi-Tier Class-Aware Token Transformer
MCAT: Visual Query-Based Localization of Standard Anatomical Clips in Fetal Ultrasound Videos Using Multi-Tier Class-Aware Token Transformer Open
Accurate standard plane acquisition in fetal ultrasound (US) videos is crucial for fetal growth assessment, anomaly detection, and adherence to clinical guidelines. However, manually selecting standard frames is time-consuming and prone to…
View article: Self-supervised Normality Learning and Divergence Vector-guided Model Merging for Zero-shot Congenital Heart Disease Detection in Fetal Ultrasound Videos
Self-supervised Normality Learning and Divergence Vector-guided Model Merging for Zero-shot Congenital Heart Disease Detection in Fetal Ultrasound Videos Open
Congenital Heart Disease (CHD) is one of the leading causes of fetal mortality, yet the scarcity of labeled CHD data and strict privacy regulations surrounding fetal ultrasound (US) imaging present significant challenges for the developmen…
View article: Expert-Agnostic Learning to Defer
Expert-Agnostic Learning to Defer Open
Learning to Defer (L2D) trains autonomous systems to handle straightforward cases while deferring uncertain ones to human experts. Recent advancements in this field have introduced methods that offer flexibility to unseen experts at test t…
View article: Fedexit - Missing Class-Agnostic Semi-Supervised Federated Learning with Extreme Imbalance Tackling Scheme
Fedexit - Missing Class-Agnostic Semi-Supervised Federated Learning with Extreme Imbalance Tackling Scheme Open
View article: FedPIA -- Permuting and Integrating Adapters leveraging Wasserstein Barycenters for Finetuning Foundation Models in Multi-Modal Federated Learning
FedPIA -- Permuting and Integrating Adapters leveraging Wasserstein Barycenters for Finetuning Foundation Models in Multi-Modal Federated Learning Open
Large Vision-Language Models typically require large text and image datasets for effective fine-tuning. However, collecting data from various sites, especially in healthcare, is challenging due to strict privacy regulations. An alternative…
View article: F$^3$OCUS -- Federated Finetuning of Vision-Language Foundation Models with Optimal Client Layer Updating Strategy via Multi-objective Meta-Heuristics
F$^3$OCUS -- Federated Finetuning of Vision-Language Foundation Models with Optimal Client Layer Updating Strategy via Multi-objective Meta-Heuristics Open
Effective training of large Vision-Language Models (VLMs) on resource-constrained client devices in Federated Learning (FL) requires the usage of parameter-efficient fine-tuning (PEFT) strategies. To this end, we demonstrate the impact of …
View article: Review of Federated Learning and Machine Learning-Based Methods for Medical Image Analysis
Review of Federated Learning and Machine Learning-Based Methods for Medical Image Analysis Open
Federated learning is an emerging technology that enables the decentralised training of machine learning-based methods for medical image analysis across multiple sites while ensuring privacy. This review paper thoroughly examines federated…
View article: Feasibility of Federated Learning from Client Databases with Different Brain Diseases and MRI Modalities
Feasibility of Federated Learning from Client Databases with Different Brain Diseases and MRI Modalities Open
Segmentation models for brain lesions in MRI are typically developed for a specific disease and trained on data with a predefined set of MRI modalities. Such models cannot segment the disease using data with a different set of MRI modaliti…
View article: The Promise of Analog Deep Learning: Recent Advances, Challenges and Opportunities
The Promise of Analog Deep Learning: Recent Advances, Challenges and Opportunities Open
Much of the present-day Artificial Intelligence (AI) utilizes artificial neural networks, which are sophisticated computational models designed to recognize patterns and solve complex problems by learning from data. However, a major bottle…
View article: Video-based Exercise Classification and Activated Muscle Group Prediction with Hybrid X3D-SlowFast Network
Video-based Exercise Classification and Activated Muscle Group Prediction with Hybrid X3D-SlowFast Network Open
This paper introduces a simple yet effective strategy for exercise classification and muscle group activation prediction (MGAP). These tasks have significant implications for personal fitness, facilitating more affordable, accessible, safe…
View article: Examining Modality Incongruity in Multimodal Federated Learning for Medical Vision and Language-based Disease Detection
Examining Modality Incongruity in Multimodal Federated Learning for Medical Vision and Language-based Disease Detection Open
Multimodal Federated Learning (MMFL) utilizes multiple modalities in each client to build a more powerful Federated Learning (FL) model than its unimodal counterpart. However, the impact of missing modality in different clients, also calle…
View article: Investigating YOLO Models Towards Outdoor Obstacle Detection For Visually Impaired People
Investigating YOLO Models Towards Outdoor Obstacle Detection For Visually Impaired People Open
The utilization of deep learning-based object detection is an effective approach to assist visually impaired individuals in avoiding obstacles. In this paper, we implemented seven different YOLO object detection models viz ., YOLO-NAS (sma…
View article: Forecasting Lithium-Ion Battery Longevity with Limited Data Availability: Benchmarking Different Machine Learning Algorithms
Forecasting Lithium-Ion Battery Longevity with Limited Data Availability: Benchmarking Different Machine Learning Algorithms Open
As the use of Lithium-ion batteries continues to grow, it becomes increasingly important to be able to predict their remaining useful life. This work aims to compare the relative performance of different machine learning algorithms, both t…
View article: Investigating YOLO Models Towards Outdoor Obstacle Detection For Visually Impaired People
Investigating YOLO Models Towards Outdoor Obstacle Detection For Visually Impaired People Open
The utilization of deep learning-based object detection is an effective approach to assist visually impaired individuals in avoiding obstacles. In this paper, we implemented seven different YOLO object detection models \textit{viz}., YOLO-…
View article: Dual Conditioned Diffusion Models for Out-Of-Distribution Detection: Application to Fetal Ultrasound Videos
Dual Conditioned Diffusion Models for Out-Of-Distribution Detection: Application to Fetal Ultrasound Videos Open
Out-of-distribution (OOD) detection is essential to improve the reliability of machine learning models by detecting samples that do not belong to the training distribution. Detecting OOD samples effectively in certain tasks can pose a chal…
View article: How Hard Is Squash? -- Towards Information Theoretic Analysis of Motor Behavior in Squash
How Hard Is Squash? -- Towards Information Theoretic Analysis of Motor Behavior in Squash Open
Fitts' law has been widely employed as a research method for analyzing tasks within the domain of Human-Computer Interaction (HCI). However, its application to non-computer tasks has remained limited. This study aims to extend the applicat…
View article: Rethinking Semi-Supervised Federated Learning: How to co-train fully-labeled and fully-unlabeled client imaging data
Rethinking Semi-Supervised Federated Learning: How to co-train fully-labeled and fully-unlabeled client imaging data Open
The most challenging, yet practical, setting of semi-supervised federated learning (SSFL) is where a few clients have fully labeled data whereas the other clients have fully unlabeled data. This is particularly common in healthcare setting…
View article: Post-Deployment Adaptation with Access to Source Data via Federated Learning and Source-Target Remote Gradient Alignment
Post-Deployment Adaptation with Access to Source Data via Federated Learning and Source-Target Remote Gradient Alignment Open
Deployment of Deep Neural Networks in medical imaging is hindered by distribution shift between training data and data processed after deployment, causing performance degradation. Post-Deployment Adaptation (PDA) addresses this by tailorin…
View article: Forecasting Lithium-Ion Battery Longevity with Limited Data Availability: Benchmarking Different Machine Learning Algorithms
Forecasting Lithium-Ion Battery Longevity with Limited Data Availability: Benchmarking Different Machine Learning Algorithms Open
View article: SPEAK WITH YOUR HANDS Using Continuous Hand Gestures to control Articulatory Speech Synthesizer
SPEAK WITH YOUR HANDS Using Continuous Hand Gestures to control Articulatory Speech Synthesizer Open
This work presents our advancements in controlling an articulatory speech synthesis engine, \textit{viz.}, Pink Trombone, with hand gestures. Our interface translates continuous finger movements and wrist flexion into continuous speech usi…
View article: Moves made easy : deep learning-based reduction of human motor control efforts leveraging categorical perceptual constraint
Moves made easy : deep learning-based reduction of human motor control efforts leveraging categorical perceptual constraint Open
The human speech motor control system takes advantage of the constraints in categorical speech perception space to reduce the index of difficulty of articulatory tasks. Taking this for inspiration, we introduce a perceptual mapping from sp…
View article: Learning Joint Articulatory-Acoustic Representations with Normalizing Flows
Learning Joint Articulatory-Acoustic Representations with Normalizing Flows Open
The articulatory geometric configurations of the vocal tract and the acoustic properties of the resultant speech sound are considered to have a strong causal relationship. This paper aims at finding a joint latent representation between th…
View article: Ultra2Speech -- A Deep Learning Framework for Formant Frequency\n Estimation and Tracking from Ultrasound Tongue Images
Ultra2Speech -- A Deep Learning Framework for Formant Frequency\n Estimation and Tracking from Ultrasound Tongue Images Open
Thousands of individuals need surgical removal of their larynx due to\ncritical diseases every year and therefore, require an alternative form of\ncommunication to articulate speech sounds after the loss of their voice box.\nThis work addr…
View article: Learning Joint Articulatory-Acoustic Representations with Normalizing\n Flows
Learning Joint Articulatory-Acoustic Representations with Normalizing\n Flows Open
The articulatory geometric configurations of the vocal tract and the acoustic\nproperties of the resultant speech sound are considered to have a strong causal\nrelationship. This paper aims at finding a joint latent representation between\…