Publications

RAVEN: Erasing Invisible Watermarks via Novel View Synthesis

SPDMark: Selective Parameter Displacement for Robust Video Watermarking

Towards Calibrating Prompt Tuning of Vision-Language Models

MOLM: Mixture of LoRA Markers

SelfOrg: Stochastic Self-Organization in Multi-Agent Systems

LoFT: Low-Rank Adaptation That Behaves Like Full Fine-Tuning

Protecting Facial Biometrics from Malicious Generative Editing via Latent Optimization

VFace: A Training-Free Approach for Diffusion-Based Video Face Swapping

Calibration-Aware Prompt Learning for Medical Vision-Language Models

A Framework for Double-Blind Federated Adaptation of Foundation Models

Robust-LLaVA: On the Effectiveness of Large-Scale Robust Image Encoders for Multi-modal Large Language Models

Forget-MI: Machine Unlearning for Forgetting Multimodal Information in Healthcare Settings

FaceAnonyMixer: Cancelable Faces via Identity Consistent Latent Space Mixing

STEREO: A Two-Stage Framework for Adversarially Robust Concept Erasing from Text-to-Image Diffusion Models

FedSECA: Sign Election and Coordinate-wise Aggregation of Gradients for Byzantine Tolerant Federated Learning

FusedVision: A Knowledge-Infusing Approach for Practical Anomaly Detection in Real-world Surveillance Videos

Towards Evaluating the Robustness of Visual State Space Models

First-place Solution to NeurIPS 2024 Invisible Watermark Removal Challenge

FIVA: Federated Inverse Variance Averaging for Universal CT Segmentation with Uncertainty Estimation

Parameter-Efficient Multimodal Adaptation for Certified Robustness of Medical Vision-Language Models

Aequa: Fair Model Rewards in Collaborative Learning via Slimmable Networks

TrojanWave: Exploiting Prompt Learning for Stealthy Backdoor Attacks on Large Audio-Language Models

Test-time Low Rank Adaptation via Confidence Maximization for Zero-Shot Generalization

FedPeWS: Personalized Warmup via Subnetworks for Enhanced Heterogeneous Federated Learning

CYCle: Choosing Your Collaborators Wisely to Enhance Collaborative Fairness in Decentralized Learning

Chameleon: A Multimodal Learning Framework Robust to Missing Modalities

Collaborative Learning of Anomalies with Privacy (CLAP) for Unsupervised Video Anomaly Detection

Attack To Defend: Exploiting Adversarial Attacks for Detecting Poisoned Models

DiffuseMix: Label-Preserving Data Augmentation with Diffusion Models

Makeup-Guided Facial Privacy Protection via Untrained Neural Network Priors

PEMMA: Parameter-Efficient Multi-Modal Adaptation for Medical Image Segmentation

BAPLe: Backdoor Attacks on Medical Foundational Models Using Prompt Learning

PromptSmooth: Certifying Robustness of Medical Vision-Language Models via Prompt Learning

SurvRNC: Learning Ordered Representations for Survival Prediction Using Rank-N-Contrast

A Federated Learning-Friendly Approach for Parameter-Efficient Fine-Tuning of SAM in 3D Segmentation

Probing the Efficacy of Federated Parameter-Efficient Fine-Tuning of Vision Transformers for Medical Image Classification

SurvCORN: Survival Analysis with Conditional Ordinal Ranking Neural Network

Redefining Contributions: Shapley-Driven Federated Learning

Dirichlet-based Uncertainty Quantification for Personalized Federated Learning with Improved Posterior Networks

Feature Map Purification for Enhancing Adversarial Robustness of Deep Timeseries Classifiers

Multi-Attribute Vision Transformers are Efficient and Robust Learners

Introducing SDICE: An Index for Assessing Diversity of Synthetic Medical Datasets

A Coarse-to-Fine Pseudo-Labeling (C2FPL) Framework for Unsupervised Video Anomaly Detection

CaPriDe Learning: Confidential and Private Decentralized Learning Based on Encryption-Friendly Distillation Loss

CLIP2Protect: Protecting Facial Privacy Using Text-Guided Makeup via Adversarial Latent Search

Evading Forensic Classifiers with Attribute-Conditioned Adversarial Faces

Lifelong Learning of Task-Parameter Relationships for Knowledge Transfer

FLIP: Cross-domain Face Anti-spoofing with Language Guidance

FeSViBS: Federated Split Learning of Vision Transformer with Block Sampling

FedSIS: Federated Split Learning with Intermediate Representation Sampling for Privacy-preserving Generalized Face Presentation Attack Detection

On Self-Supervised Learning and Prompt Tuning of Vision Transformers for Cross-sensor Fingerprint Presentation Attack Detection

Recurrence-based Disentanglement for Detecting Adversarial Attacks on Timeseries Classifiers

Towards Building Text-To-Speech Systems for the Next Billion Users

Single-branch Network for Multimodal Training

PolyKervNets: Activation-free Neural Networks For Efficient Private Inference

A Unified Model for Face Matching and Presentation Attack Detection Using an Ensemble of Vision Transformer Features

Self-omics: A Self-supervised Learning Framework for Multi-omics Cancer Data

Byzantine-Tolerant Methods for Distributed Variational Inequalities

DCTM: Dilated Convolutional Transformer Model for Multimodal Engagement Estimation in Conversation

How to Democratise and Protect AI: Fair and Differentially Private Decentralised Deep Learning

Self-Ensembling Vision Transformer (SEViT) for Robust Medical Image Classification

Suppressing Poisoning Attacks on Federated Learning for Medical Imaging

On the Importance of Image Encoding in Automated Chest X-Ray Report Generation

On Demographic Bias in Fingerprint Recognition

Hate-CLIPper: Multimodal Hateful Meme Classification based on Cross-modal Interaction of CLIP Features

MuCoT: Multilingual Contrastive Training for Question-Answering in Low-resource Languages

SubOmiEmbed: Self-supervised Representation Learning of Multi-omics Data for Cancer Type Classification

Learning an Ensemble of Deep Fingerprint Representations

FHE-Friendly Distillation of Decision Tree Ensembles for Efficient Encrypted Inference

CryptInfer — Enabling Practical Inference on Encrypted Skin Lesion Images for Melanoma Detection

Optimizing Homomorphic Encryption Based Secure Image Analytics

Continual Domain Incremental Learning for Chest X-ray Classification in Low-Resource Clinical Settings

Efficient CNN Building Blocks for Encrypted Data

Cancelable Biometrics Vault: A Secure Key-Binding Biometric Cryptosystem Based on Chaffing and Winnowing