Nazanin Rahnavard

Nazanin Rahnavard

Auto-generated profile, based on publicly available information

Nazanin Rahnavard

University of Central Florida

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
14 August 2024 in IEEE Transactions on Wireless Communications
Verified
Journal Article
Closed access
KVKatarina Vuckovic
Katarina Vuckovic
SMSaba Mohammadhosseini
Saba Mohammadhosseini
FHFarzam Hejazi
Farzam Hejazi
NRNazanin Rahnavard
Nazanin Rahnavard
This paper presents a new method for user localization in single-site massive Multiple-Input-Multiple-Output (MIMO) systems, which circumvents the need for large labeled datasets typically required for training data-driven models. Instead, the proposed model utilizes a limited set of geo-tagged Channel State Information (CSI) samples for training. The approach combines a Fully-Connected Auto-Encoder (FC-AE) with a Gaussian Process Regression (GPR) model. The GPR model is efficient, as it requires only a minimal amount of labeled data for training, although it presents challenges in computational complexity. To address this complexity, the FC-AE is introduced, which encodes the Angle-Delay Profile (ADP) transformation of the CSI data. The training dataset for the FC-AE is crafted by employing data augmentation techniques on a small collection of unlabeled data. The simulation results demonstrate that FC-AE is scenario-independent and adaptable to new scenarios with similar ADP characteristics. Additionally, our FC-AE-GPR model surpasses the performance of the Convolutional Neural Network model and the non-parametric grid search method when provided with limited labeled data, applicable in both indoor and outdoor settings.

2 December 2024
Verified
Conference Paper
Open access
NKNazmul Karim
Nazmul Karim
AAAbdullah Al Arafat
Abdullah Al Arafat
ARAdnan Siraj Rakin
Adnan Siraj Rakin
ZGZhishan Guo
Zhishan Guo
NRNazanin Rahnavard
Nazanin Rahnavard
Studies on backdoor attacks in recent years suggest that an adversary can compromise the integrity of a deep neural network (DNN) by manipulating a small set of training samples. Our analysis shows that such manipulation can make the backdoor model converge to a bad local minima, i.e., sharper minima as compared to a benign model. Intuitively, the backdoor can be purified by re-optimizing the model to smoother minima. However, a naïve adoption of any optimization targeting smoother minima can lead to sub-optimal purification techniques hampering the clean test accuracy. Hence, to effectively obtain such re-optimization, inspired by our novel perspective establishing the connection between backdoor removal and loss smoothness, we propose <u>F</u>isher <u>I</u>nformation guided <u>P</u>urification (FIP), a novel backdoor purification framework. Proposed FIP consists of a couple of novel regularizers that aid the model in suppressing the backdoor effects and retaining the acquired knowledge of clean data distribution throughout the backdoor removal procedure through exploiting the knowledge of Fisher Information Matrix (FIM). In addition, we introduce an efficient variant of FIP, dubbed as Fast FIP, which reduces the number of tunable parameters significantly and obtains an impressive runtime gain of almost 5×. Extensive experiments show that the proposed method achieves state-of-the-art (SOTA) performance on a wide range of backdoor defense benchmarks: 5 different tasks---Image Recognition, Object Detection, Video Action Recognition, 3D point Cloud, Language Generation ; 11 different datasets including ImageNet, PASCAL VOC, UCF101 ; diverse model architectures spanning both CNN and vision transformer; 14 different backdoor attacks, e.g., Dynamic, WaNet, LIRA, ISSBA, etc. Our code is available in this https://github.com/nazmul-karim170/FIP-Fisher-Backdoor-Removal GitHub Repository.

26 October 2024
Verified
Book Chapter
Closed access
PDF file
NKNazmul Karim
Nazmul Karim
AAAbdullah Al Arafat
Abdullah Al Arafat
UKUmar Khalid
Umar Khalid
ZGZhishan Guo
Zhishan Guo
NRNazanin Rahnavard
Nazanin Rahnavard

17 July 2024 in arXiv
Verified
Preprint
Open access
PDF file
NKNazmul Karim
Nazmul Karim
AAAbdullah Al Arafat
Abdullah Al Arafat
UKUmar Khalid
Umar Khalid
ZGZhishan Guo
Zhishan Guo
NRNazanin Rahnavard
Nazanin Rahnavard
Recent studies have revealed the vulnerability of deep neural networks (DNNs) to various backdoor attacks, where the behavior of DNNs can be compromised by utilizing certain types of triggers or poisoning mechanisms. State-of-the-art (SOTA) defenses employ too-sophisticated mechanisms that require either a computationally expensive adversarial search module for reverse-engineering the trigger distribution or an over-sensitive hyper-parameter selection module. Moreover, they offer sub-par performance in challenging scenarios, e.g., limited validation data and strong attacks. In this paper, we propose Neural mask Fine-Tuning (NFT) with an aim to optimally re-organize the neuron activities in a way that the effect of the backdoor is removed. Utilizing a simple data augmentation like MixUp, NFT relaxes the trigger synthesis process and eliminates the requirement of the adversarial search module. Our study further reveals that direct weight fine-tuning under limited validation data results in poor post-purification clean test accuracy, primarily due to overfitting issue. To overcome this, we propose to fine-tune neural masks instead of model weights. In addition, a mask regularizer has been devised to further mitigate the model drift during the purification process. The distinct characteristics of NFT render it highly efficient in both runtime and sample usage, as it can remove the backdoor even when a single sample is available from each class. We validate the effectiveness of NFT through extensive experiments covering the tasks of image classification, object detection, video action recognition, 3D point cloud, and natural language processing. We evaluate our method against 14 different attacks (LIRA, WaNet, etc.) on 11 benchmark data sets such as ImageNet, UCF101, Pascal VOC, ModelNet, OpenSubtitles2012, etc.