arXiv subjects

Nazanin Rahnavard

Publications and source records attributed to Nazanin Rahnavard.

At least 19 recordsLinked to original sources

arXiv2025

EVLM: Self-Reflective Multimodal Reasoning for Cross-Dimensional Visual Editing

Editing complex visual content from ambiguous or partially specified instructions remains a core challenge in vision-language modeling. Existing models can contextualize content but often fail to infer the underlying intent within a reference image or scene, leading to inconsistent or misaligned edits. We introduce the Editing Vision-Language Model (EVLM), a system that interprets ambiguous instructions in conjunction with reference visuals to produce precise, context-aware editing prompts. EVLM's key innovation is a reflective reasoning framework that translates subjective user intent into structured, actionable outputs by aligning with human-rated rationales through Reflection-Aware KL-Divergence Target Optimization (RKTO). By combining Chain-of-Thought (CoT) reasoning with RKTO alignment, EVLM captures fine-grained editing preferences without relying on binary supervision. Trained on a dataset of 30,000 CoT examples with human-annotated rationale quality, EVLM achieves substantial gains in alignment with human intent. Experiments across image, video, 3D, and 4D editing tasks show that EVLM generates coherent and high-quality instructions, providing a scalable foundation for multimodal editing and reasoning.

cs.CV↗

arXiv2025

On Transfer-based Universal Attacks in Pure Black-box Setting

Despite their impressive performance, deep visual models are susceptible to transferable black-box adversarial attacks. Principally, these attacks craft perturbations in a target model-agnostic manner. However, surprisingly, we find that existing methods in this domain inadvertently take help from various priors that violate the black-box assumption such as the availability of the dataset used to train the target model, and the knowledge of the number of classes in the target model. Consequently, the literature fails to articulate the true potency of transferable black-box attacks. We provide an empirical study of these biases and propose a framework that aids in a prior-free transparent study of this paradigm. Using our framework, we analyze the role of prior knowledge of the target model data and number of classes in attack performance. We also provide several interesting insights based on our analysis, and demonstrate that priors cause overestimation in transferability scores. Finally, we extend our framework to query-based attacks. This extension inspires a novel image-blending technique to prepare data for effective surrogate model training.

cs.LG↗

arXiv2024

Fisher Information guided Purification against Backdoor Attacks

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 Fisher Information guided Purification (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 $5\times$. 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.

cs.CV↗

arXiv2024

Augmented Neural Fine-Tuning for Efficient Backdoor Purification

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.

cs.CV↗

arXiv2023

SAVE: Spectral-Shift-Aware Adaptation of Image Diffusion Models for Text-driven Video Editing

Text-to-Image (T2I) diffusion models have achieved remarkable success in synthesizing high-quality images conditioned on text prompts. Recent methods have tried to replicate the success by either training text-to-video (T2V) models on a very large number of text-video pairs or adapting T2I models on text-video pairs independently. Although the latter is computationally less expensive, it still takes a significant amount of time for per-video adaption. To address this issue, we propose SAVE, a novel spectral-shift-aware adaptation framework, in which we fine-tune the spectral shift of the parameter space instead of the parameters themselves. Specifically, we take the spectral decomposition of the pre-trained T2I weights and only update the singular values while freezing the corresponding singular vectors. In addition, we introduce a spectral shift regularizer aimed at placing tighter constraints on larger singular values compared to smaller ones. This form of regularization enables the model to grasp finer details within the video that align with the provided textual descriptions. We also offer theoretical justification for our proposed regularization technique. Since we are only dealing with spectral shifts, the proposed method reduces the adaptation time significantly (approx. 10 times) and has fewer resource constraints for training. Such attributes posit SAVE to be more suitable for real-world applications, e.g. editing undesirable content during video streaming. We validate the effectiveness of SAVE with an extensive experimental evaluation under different settings, e.g. style transfer, object replacement, privacy preservation, etc.

cs.CV↗

arXiv2023

CSI-Based Data-driven Localization Frameworking using Small-scale Training Datasets in Single-site MIMO Systems

This work presents a date-driven user localization framework for single-site massive Multiple-Input-Multiple-Output (MIMO) systems. The framework is trained on a geo-tagged Channel State Information (CSI) dataset. Unlike the state-of-the-art Convolutional Neural Network (CNN) models, which require large training datasets to perform well, our method is specifically designed to operate with small-scale training datasets. This makes our approach more practical for real-world scenarios, where collecting a large amount of data can be challenging. Our proposed FC-AE-GPR framework combines two components: a Fully-Connected Auto-Encoder (FC-AE) and a Gaussian Process Regression (GPR) model. Our results show that the GPR model outperforms the CNN model when presented with small training datasets. However, the training complexity of GPR models can become an issue when the input sample size is large. To address this, we propose using the FC-AE to reduce the sample size by encoding the CSI before training the GPR model. Although the FC-AE model may require a larger training dataset initially, we demonstrate that the FC-AE is scenario independent. This means that it can be utilized in new and unseen scenarios without prior retraining. Therefore, adapting the FC-AE-GPR model to a new scenario requires only retraining the GPR model with a small training dataset.

eess.SP↗

arXiv2023

C-SFDA: A Curriculum Learning Aided Self-Training Framework for Efficient Source Free Domain Adaptation

Unsupervised domain adaptation (UDA) approaches focus on adapting models trained on a labeled source domain to an unlabeled target domain. UDA methods have a strong assumption that the source data is accessible during adaptation, which may not be feasible in many real-world scenarios due to privacy concerns and resource constraints of devices. In this regard, source-free domain adaptation (SFDA) excels as access to source data is no longer required during adaptation. Recent state-of-the-art (SOTA) methods on SFDA mostly focus on pseudo-label refinement based self-training which generally suffers from two issues: i) inevitable occurrence of noisy pseudo-labels that could lead to early training time memorization, ii) refinement process requires maintaining a memory bank which creates a significant burden in resource constraint scenarios. To address these concerns, we propose C-SFDA, a curriculum learning aided self-training framework for SFDA that adapts efficiently and reliably to changes across domains based on selective pseudo-labeling. Specifically, we employ a curriculum learning scheme to promote learning from a restricted amount of pseudo labels selected based on their reliabilities. This simple yet effective step successfully prevents label noise propagation during different stages of adaptation and eliminates the need for costly memory-bank based label refinement. Our extensive experimental evaluations on both image recognition and semantic segmentation tasks confirm the effectiveness of our method. C-SFDA is readily applicable to online test-time domain adaptation and also outperforms previous SOTA methods in this task.

cs.CV↗

arXiv2022

RODD: A Self-Supervised Approach for Robust Out-of-Distribution Detection

Recent studies have addressed the concern of detecting and rejecting the out-of-distribution (OOD) samples as a major challenge in the safe deployment of deep learning (DL) models. It is desired that the DL model should only be confident about the in-distribution (ID) data which reinforces the driving principle of the OOD detection. In this paper, we propose a simple yet effective generalized OOD detection method independent of out-of-distribution datasets. Our approach relies on self-supervised feature learning of the training samples, where the embeddings lie on a compact low-dimensional space. Motivated by the recent studies that show self-supervised adversarial contrastive learning helps robustify the model, we empirically show that a pre-trained model with self-supervised contrastive learning yields a better model for uni-dimensional feature learning in the latent space. The method proposed in this work referred to as RODD outperforms SOTA detection performance on an extensive suite of benchmark datasets on OOD detection tasks. On the CIFAR-100 benchmarks, RODD achieves a 26.97 $\%$ lower false-positive rate (FPR@95) compared to SOTA methods.

cs.CV↗

arXiv2022

UNICON: Combating Label Noise Through Uniform Selection and Contrastive Learning

Supervised deep learning methods require a large repository of annotated data; hence, label noise is inevitable. Training with such noisy data negatively impacts the generalization performance of deep neural networks. To combat label noise, recent state-of-the-art methods employ some sort of sample selection mechanism to select a possibly clean subset of data. Next, an off-the-shelf semi-supervised learning method is used for training where rejected samples are treated as unlabeled data. Our comprehensive analysis shows that current selection methods disproportionately select samples from easy (fast learnable) classes while rejecting those from relatively harder ones. This creates class imbalance in the selected clean set and in turn, deteriorates performance under high label noise. In this work, we propose UNICON, a simple yet effective sample selection method which is robust to high label noise. To address the disproportionate selection of easy and hard samples, we introduce a Jensen-Shannon divergence based uniform selection mechanism which does not require any probabilistic modeling and hyperparameter tuning. We complement our selection method with contrastive learning to further combat the memorization of noisy labels. Extensive experimentation on multiple benchmark datasets demonstrates the effectiveness of UNICON; we obtain an 11.4% improvement over the current state-of-the-art on CIFAR100 dataset with a 90% noise rate. Our code is publicly available

cs.CV↗

arXiv2022

CNLL: A Semi-supervised Approach For Continual Noisy Label Learning

The task of continual learning requires careful design of algorithms that can tackle catastrophic forgetting. However, the noisy label, which is inevitable in a real-world scenario, seems to exacerbate the situation. While very few studies have addressed the issue of continual learning under noisy labels, long training time and complicated training schemes limit their applications in most cases. In contrast, we propose a simple purification technique to effectively cleanse the online data stream that is both cost-effective and more accurate. After purification, we perform fine-tuning in a semi-supervised fashion that ensures the participation of all available samples. Training in this fashion helps us learn a better representation that results in state-of-the-art (SOTA) performance. Through extensive experimentation on 3 benchmark datasets, MNIST, CIFAR10 and CIFAR100, we show the effectiveness of our proposed approach. We achieve a 24.8% performance gain for CIFAR10 with 20% noise over previous SOTA methods. Our code is publicly available.

cs.CV↗

arXiv2022

CoDGraD: A Code-based Distributed Gradient Descent Scheme for Decentralized Convex Optimization

In this paper, we consider a large network containing many regions such that each region is equipped with a worker with some data processing and communication capability. For such a network, some workers may become stragglers due to the failure or heavy delay on computing or communicating. To resolve the above straggling problem, a coded scheme that introduces certain redundancy for every worker was recently proposed, and a gradient coding paradigm was developed to solve convex optimization problems when the network has a centralized fusion center. In this paper, we propose an iterative distributed algorithm, referred as Code-Based Distributed Gradient Descent algorithm (CoDGraD), to solve convex optimization problems over distributed networks. In each iteration of the proposed algorithm, an active worker shares the coded local gradient and approximated solution of the convex optimization problem with non-straggling workers at the adjacent regions only. In this paper, we also provide the consensus and convergence analysis for the CoDGraD algorithm and we demonstrate its performance via numerical simulations.

eess.SY↗

arXiv2022

RF Signal Transformation and Classification using Deep Neural Networks

Deep neural networks (DNNs) designed for computer vision and natural language processing tasks cannot be directly applied to the radio frequency (RF) datasets. To address this challenge, we propose to convert the raw RF data to data types that are suitable for off-the-shelf DNNs by introducing a convolutional transform technique. In addition, we propose a simple 5-layer convolutional neural network architecture (CONV-5) that can operate with raw RF I/Q data without any transformation. Further, we put forward an RF dataset, referred to as RF1024, to facilitate future RF research. RF1024 consists of 8 different RF modulation classes with each class having 1000/200 training/test samples. Each sample of the RF1024 dataset contains 1024 complex I/Q values. Lastly, the experiments are performed on the RadioML2016 and RF1024 datasets to demonstrate the improved classification performance.

eess.SP↗

arXiv2021

MAP-CSI: Single-site Map-Assisted Localization Using Massive MIMO CSI

This paper presents a new map-assisted localization approach utilizing Chanel State Information (CSI) in Massive Multiple-Input Multiple-Output (MIMO) systems. Map-assisted localization is an environment-aware approach in which the communication system has information regarding the surrounding environment. By combining radio frequency ray tracing parameters of the multipath components (MPC) with the environment map, it is possible to accomplish localization. Unfortunately, in real-world scenarios, ray tracing parameters are typically not explicitly available. Thus, additional complexity is added at a base station to obtain this information. On the other hand, CSI is a common communication parameter, usually estimated for any communication channel. In this work, we leverage the already available CSI data to propose a novel map-assisted CSI localization approach, referred to as MAP-CSI. We show that Angle-of-Departure (AoD) and Time-of-Arrival (ToA) can be extracted from CSI and then be used in combination with the environment map to localize the user. We perform simulations on a public MIMO dataset and show that our method works for both line-of-sight (LOS) and non-line-of-sight (NLOS) scenarios. We compare our method to the state-of-the-art (SoA) method that uses the ray tracing data. Using MAP-CSI, we accomplish an average localization error of 1.8 m in LOS and 2.8 m in mixed (combination of LOS and NLOS samples) scenarios. On the other hand, SoA ray tracing has an average error of 1.0 m and 2.2 m, respectively, but requires explicit AoD and ToA information to perform the localization task.

eess.SP↗

arXiv2021

Straggler-Robust Distributed Optimization in Parameter-Server Networks

Optimization in distributed networks plays a central role in almost all distributed machine learning problems. In principle, the use of distributed task allocation has reduced the computational time, allowing better response rates and higher data reliability. However, for these computational algorithms to run effectively in complex distributed systems, the algorithms ought to compensate for communication asynchrony, network node failures and delays known as stragglers. These issues can change the effective connection topology of the network, which may vary over time, thus hindering the optimization process. In this paper, we propose a new distributed unconstrained optimization algorithm for minimizing a convex function which is adaptable to a parameter server network. In particular, the network worker nodes solve their local optimization problems, allowing the computation of their local coded gradients, which will be sent to different server nodes. Then within this parameter server platform each server node aggregates its communicated local gradients, allowing convergence to the desired optimizer. This algorithm is robust to network s worker node failures, disconnection, or delaying nodes known as stragglers. One way to overcome the straggler problem is to allow coding over the network. We further extend this coding framework to enhance the convergence of the proposed algorithm under such varying network topologies. By using coding and utilizing evaluations of gradients of uniformly bounded delay we further enhance the proposed algorithm performance. Finally, we implement the proposed scheme in MATLAB and provide comparative results demonstrating the effectiveness of the proposed framework

math.OC↗

arXiv2021

Spectrum Shaping For Multiple Link Discovery in 6G THz Systems

This paper presents a novel antenna configuration to measure directions of multiple signal sources at the receiver in a THz mobile network via a single channel measurement. Directional communication is an intrinsic attribute of THz wireless networks and the knowledge of direction should be harvested continuously to maintain link quality. Direction discovery can potentially impose an immense burden on the network that limits its communication capacity exceedingly. To utterly mitigate direction discovery overhead, we propose a novel technique called spectrum shaping capable of measuring direction, power, and relative distance of propagation paths via a single measurement. We demonstrate that the proposed technique is also able to measure the transmitter antenna orientation. We evaluate the performance of the proposed design in several scenarios and show that the introduced technique performs similar to a large array of antennas while attaining a much simpler hardware architecture. Results show that the spectrum shaping with only two antennas placed 0.5 mm, 5 mm, and 1 cm apart performs direction of arrival estimation similar to a much more complex uniform linear array equipped with 7, 60, and 120 antennas, respectively.

eess.SP↗

arXiv2021

Phase Spectrometry For High Precision mm-Wave DoA Estimation In 5G Systems

In this paper, we introduce a direction of arrival (DoA) estimation method based on a technique named phase spectrometry (PS) that is mainly suitable for mm-Wave and Tera-hertz applications as an alternative for DoA estimation using antenna arrays. PS is a conventional technique in optics to measure phase difference between two waves at different frequencies of the spectrum. Here we adapt PS for the same purpose in the radio frequency band. We show that we can emulate a large array exploiting only two antennas. To this end, we measure phase difference between the two antennas for different frequencies using PS. Consequently, we demonstrate that we can radically reduce the complexity of the receiver required for DoA estimation employing PS. We consider two different schemes for implementation of PS: via a long wave-guide and frequency code-book. We show that using a frequency code-book, higher processing gain can be achieved. Moreover, we introduce three PS architectures: for device to device DoA estimation, for base-station in uplink scenario and an ultra-fast DoA estimation technique mainly for radar and aerial and satellite communications. Simulation and analytical results show that, PS is capable of detecting and discriminating between multiple incoming signals with different DoAs. Moreover, our results also show that, the angular resolution of PS depends on the distance between the two antennas and the band-width of the frequency code-book. Finally, the performance of PS is compared with a uniform linear array (ULA) and it is shown that PS can perform the same, with a much less complex receiver, and without the prerequisite of spatial search for DoA estimation.

eess.SP↗

arXiv2021

SPI-GAN: Towards Single-Pixel Imaging through Generative Adversarial Network

Single-pixel imaging is a novel imaging scheme that has gained popularity due to its huge computational gain and potential for a low-cost alternative to imaging beyond the visible spectrum. The traditional reconstruction methods struggle to produce a clear recovery when one limits the number of illumination patterns from a spatial light modulator. As a remedy, several deep-learning-based solutions have been proposed which lack good generalization ability due to the architectural setup and loss functions. In this paper, we propose a generative adversarial network-based reconstruction framework for single-pixel imaging, referred to as SPI-GAN. Our method can reconstruct images with 17.92 dB PSNR and 0.487 SSIM, even if the sampling ratio drops to 5%. This facilitates much faster reconstruction making our method suitable for single-pixel video. Furthermore, our ResNet-like architecture for the generator leads to useful representation learning that allows us to reconstruct completely unseen objects. The experimental results demonstrate that SPI-GAN achieves significant performance gain, e.g. near 3dB PSNR gain, over the current state-of-the-art method.

cs.CV↗

arXiv2021

RL-NCS: Reinforcement learning based data-driven approach for nonuniform compressed sensing

A reinforcement-learning-based non-uniform compressed sensing (NCS) framework for time-varying signals is introduced. The proposed scheme, referred to as RL-NCS, aims to boost the performance of signal recovery through an optimal and adaptive distribution of sensing energy among two groups of coefficients of the signal, referred to as the region of interest (ROI) coefficients and non-ROI coefficients. The coefficients in ROI usually have greater importance and need to be reconstructed with higher accuracy compared to non-ROI coefficients. In order to accomplish this task, the ROI is predicted at each time step using two specific approaches. One of these approaches incorporates a long short-term memory (LSTM) network for the prediction. The other approach employs the previous ROI information for predicting the next step ROI. Using the exploration-exploitation technique, a Q-network learns to choose the best approach for designing the measurement matrix. Furthermore, a joint loss function is introduced for the efficient training of the Q-network as well as the LSTM network. The result indicates a significant performance gain for our proposed method, even for rapidly varying signals and a reduced number of measurements.

cs.LG↗