arXiv subjects

Muzammil Behzad

Publications and source records attributed to Muzammil Behzad.

At least 19 recordsLinked to original sources

arXiv2026

ROI-Gated SAHI: Content-Adaptive Slicing-Based Inference for Efficient Object Detection

Slicing-Aided Hyper Inference (SAHI) improves small object detection in high-resolution images but often spends substantial compute on background tiles. We propose region-of-interest (ROI)-Gated SAHI, an inference-time framework that introduces a lightweight proposer to localize foreground regions and restrict sliced refinement to informative areas. We evaluate the framework in two settings. On the COCO128 full split dataset comprising 128 images, static ROI-gating is slower on average than Full SAHI, achieving a speed ratio of 0.88, and yields a lower mAP@0.5 of 0.6602 compared with 0.7569 for Full SAHI. A simple adaptive routing policy with $τ=$ 0.4 educes the mean latency, achieving a slight gain of 1.02$\times$ over Full SAHI. On a three-image sparse-to-dense case study, ROI-gating achieves speedups ranging from 0.96$\times$ to 6.90$\times$ with a mean speedup of 3.41$\times$. These results show that ROI-gating is most beneficial in sparse scenes and requires policy-based routing for robust average behavior.

cs.CV↗

arXiv2026

The Professor: Multi-Teacher Unsupervised Prompt Distillation for Vision-Language Models

Prompt distillation compresses large vision-language models (VLMs) such as CLIP into lightweight student models by matching teacher predictions on unlabeled domain images. PromptKD (CVPR 2024) established this paradigm with a single PromptSRC-finetuned ViT-L/14 teacher and a ViT-B/16 student. We propose TheProfessor, a multi-teacher extension that distills from a fixed two-teacher ensemble: a domain-finetuned PromptSRC ViT-L/14 teacher and a zero-shot EVA-CLIP-L/14 teacher whose logits are pre-computed per dataset. We evaluate single-teacher PromptKD, equal-probability ensembling, and confidence-weighted ensembling on four base-to-novel datasets: Caltech-101, DTD, UCF101, and EuroSAT. In a 12-run single-seed sweep, confidence-weighted ensembling improves average HM from 87.52 to 89.28 (+1.77 points), while equal averaging improves average HM to 88.88 (+1.37 points). Gains are dataset dependent: they are negligible on Caltech-101 (+0.16 HM for confidence weighting), modest on UCF101 (+0.62), and largest on domain-shifted EuroSAT (+5.78). These results update our earlier Caltech-only analysis and show that multi-teacher prompt distillation is most useful when the second teacher contributes complementary supervision under domain shift.

cs.CV↗

arXiv2026

Hallucination in Medical Imaging AI: A Cross-Modality Analytical Framework for Taxonomy, Detection, and Mitigation under Regulatory Constraints

AI systems are being deployed across medical imaging faster than their failure modes are understood. At this point in time, the failure of greatest clinical concern is hallucination: clinically plausible but factually incorrect outputs, including fabricated anatomical structures, missed findings, incorrect laterality, and invented measurements in generated reports, with direct consequences, for example, for biopsy decisions, staging, and treatment planning. This structured narrative synthesizes peer-reviewed studies, benchmark datasets, and FDA regulatory guidance across five imaging modalities to produce a cross-modality analysis of hallucination taxonomy, etiology, detection, and mitigation. Specifically, we address three questions in this study: (1) how can existing taxonomies be unified across modalities?, (2) how do medical-specialized foundation models hallucinate less than general-purpose ones?, and (3) which mitigation strategies are effective and compatible with FDA lifecycle oversight? We note that three taxonomic frameworks together cover the imaging pipeline in a way no single framework does alone. We also highlight that general-purpose foundation models outperform medical-specialized models on hallucination-specific benchmarks, indicating that narrow domain fine-tuning can introduce overfitting-induced confabulation. At the same time, the oversight of radiologists remains essential; for instance, a very high percentage of of AI-generated flags required expert correction before clinical use. Physics-informed architectural constraints, Chain-of-Thought prompting, and human-in-the-loop safeguards each address different failure modes and is effective when combined. All findings are mapped to the FDA's Total Product Lifecycle and Predetermined Change Control Plan frameworks, which treat hallucination management as a lifecycle obligation rather than a pre-deployment checklist.

cs.AI↗

arXiv2026

Spatio-Semantic Expert Routing Architecture with Mixture-of-Experts for Referring Image Segmentation

Referring image segmentation aims to produce a pixel-level mask for the image region described by a natural-language expression. Although pretrained vision-language models have improved semantic grounding, many existing methods still rely on uniform refinement strategies that do not fully match the diverse reasoning requirements of referring expressions. Because of this mismatch, predictions often contain fragmented regions, inaccurate boundaries, or even the wrong object, especially when pretrained backbones are frozen for computational efficiency. To address these limitations, we propose SERA, a Spatio-Semantic Expert Routing Architecture for referring image segmentation. SERA introduces lightweight, expression-aware expert refinement at two complementary stages within a vision-language framework. First, we design SERA-Adapter, which inserts an expression-conditioned adapter into selected backbone blocks to improve spatial coherence and boundary precision through expert-guided refinement and cross-modal attention. We then introduce SERA-Fusion, which strengthens intermediate visual representations by reshaping token features into spatial grids and applying geometry-preserving expert transformations before multimodal interaction. In addition, a lightweight routing mechanism adaptively weights expert contributions while remaining compatible with pretrained representations. To make this routing stable under frozen encoders, SERA uses a parameter-efficient tuning strategy that updates only normalization and bias terms, affecting less than 1% of the backbone parameters. Experiments on standard referring image segmentation benchmarks show that SERA consistently outperforms strong baselines, with especially clear gains on expressions that require accurate spatial localization and precise boundary delineation.

cs.CV↗

arXiv2026

XAI-CLIP: ROI-Guided Perturbation Framework for Explainable Medical Image Segmentation in Multimodal Vision-Language Models

Medical image segmentation is a critical component of clinical workflows, enabling accurate diagnosis, treatment planning, and disease monitoring. However, despite the superior performance of transformer-based models over convolutional architectures, their limited interpretability remains a major obstacle to clinical trust and deployment. Existing explainable artificial intelligence (XAI) techniques, including gradient-based saliency methods and perturbation-based approaches, are often computationally expensive, require numerous forward passes, and frequently produce noisy or anatomically irrelevant explanations. To address these limitations, we propose XAI-CLIP, an ROI-guided perturbation framework that leverages multimodal vision-language model embeddings to localize clinically meaningful anatomical regions and guide the explanation process. By integrating language-informed region localization with medical image segmentation and applying targeted, region-aware perturbations, the proposed method generates clearer, boundary-aware saliency maps while substantially reducing computational overhead. Experiments conducted on the FLARE22 and CHAOS datasets demonstrate that XAI-CLIP achieves up to a 60\% reduction in runtime, a 44.6\% improvement in dice score, and a 96.7\% increase in Intersection-over-Union for occlusion-based explanations compared to conventional perturbation methods. Qualitative results further confirm cleaner and more anatomically consistent attribution maps with fewer artifacts, highlighting that the incorporation of multimodal vision-language representations into perturbation-based XAI frameworks significantly enhances both interpretability and efficiency, thereby enabling transparent and clinically deployable medical image segmentation systems.

cs.CV↗

arXiv2025

SwinTF3D: A Lightweight Multimodal Fusion Approach for Text-Guided 3D Medical Image Segmentation

The recent integration of artificial intelligence into medical imaging has driven remarkable advances in automated organ segmentation. However, most existing 3D segmentation frameworks rely exclusively on visual learning from large annotated datasets restricting their adaptability to new domains and clinical tasks. The lack of semantic understanding in these models makes them ineffective in addressing flexible, user-defined segmentation objectives. To overcome these limitations, we propose SwinTF3D, a lightweight multimodal fusion approach that unifies visual and linguistic representations for text-guided 3D medical image segmentation. The model employs a transformer-based visual encoder to extract volumetric features and integrates them with a compact text encoder via an efficient fusion mechanism. This design allows the system to understand natural-language prompts and correctly align semantic cues with their corresponding spatial structures in medical volumes, while producing accurate, context-aware segmentation results with low computational overhead. Extensive experiments on the BTCV dataset demonstrate that SwinTF3D achieves competitive Dice and IoU scores across multiple organs, despite its compact architecture. The model generalizes well to unseen data and offers significant efficiency gains compared to conventional transformer-based segmentation networks. Bridging visual perception with linguistic understanding, SwinTF3D establishes a practical and interpretable paradigm for interactive, text-driven 3D medical image segmentation, opening perspectives for more adaptive and resource-efficient solutions in clinical imaging.

cs.CV↗

arXiv2025

Attack-Aware Deepfake Detection under Counter-Forensic Manipulations

This work presents an attack-aware deepfake and image-forensics detector designed for robustness, well-calibrated probabilities, and transparent evidence under realistic deployment conditions. The method combines red-team training with randomized test-time defense in a two-stream architecture, where one stream encodes semantic content using a pretrained backbone and the other extracts forensic residuals, fused via a lightweight residual adapter for classification, while a shallow Feature Pyramid Network style head produces tamper heatmaps under weak supervision. Red-team training applies worst-of-K counter-forensics per batch, including JPEG realign and recompress, resampling warps, denoise-to-regrain operations, seam smoothing, small color and gamma shifts, and social-app transcodes, while test-time defense injects low-cost jitters such as resize and crop phase changes, mild gamma variation, and JPEG phase shifts with aggregated predictions. Heatmaps are guided to concentrate within face regions using face-box masks without strict pixel-level annotations. Evaluation on existing benchmarks, including standard deepfake datasets and a surveillance-style split with low light and heavy compression, reports clean and attacked performance, AUC, worst-case accuracy, reliability, abstention quality, and weak-localization scores. Results demonstrate near-perfect ranking across attacks, low calibration error, minimal abstention risk, and controlled degradation under regrain, establishing a modular, data-efficient, and practically deployable baseline for attack-aware detection with calibrated probabilities and actionable heatmaps.

cs.CV↗

arXiv2025

AquaDiff: Diffusion-Based Underwater Image Enhancement for Addressing Color Distortion

Underwater images are severely degraded by wavelength-dependent light absorption and scattering, resulting in color distortion, low contrast, and loss of fine details that hinder vision-based underwater applications. To address these challenges, we propose AquaDiff, a diffusion-based underwater image enhancement framework designed to correct chromatic distortions while preserving structural and perceptual fidelity. AquaDiff integrates a chromatic prior-guided color compensation strategy with a conditional diffusion process, where cross-attention dynamically fuses degraded inputs and noisy latent states at each denoising step. An enhanced denoising backbone with residual dense blocks and multi-resolution attention captures both global color context and local details. Furthermore, a novel cross-domain consistency loss jointly enforces pixel-level accuracy, perceptual similarity, structural integrity, and frequency-domain fidelity. Extensive experiments on multiple challenging underwater benchmarks demonstrate that AquaDiff provides good results as compared to the state-of-the-art traditional, CNN-, GAN-, and diffusion-based methods, achieving superior color correction and competitive overall image quality across diverse underwater conditions.

cs.CV↗

arXiv2025

From Graphs to Gates: DNS-HyXNet, A Lightweight and Deployable Sequential Model for Real-Time DNS Tunnel Detection

Domain Name System (DNS) tunneling remains a covert channel for data exfiltration and command-and-control communication. Although graph-based methods such as GraphTunnel achieve strong accuracy, they introduce significant latency and computational overhead due to recursive parsing and graph construction, limiting their suitability for real-time deployment. This work presents DNS-HyXNet, a lightweight extended Long Short-Term Memory (xLSTM) hybrid framework designed for efficient sequence-based DNS tunnel detection. DNS-HyXNet integrates tokenized domain embeddings with normalized numerical DNS features and processes them through a two-layer xLSTM network that directly learns temporal dependencies from packet sequences, eliminating the need for graph reconstruction and enabling single-stage multi-class classification. The model was trained and evaluated on two public benchmark datasets with carefully tuned hyperparameters to ensure low memory consumption and fast inference. Across all experimental splits of the DNS-Tunnel-Datasets, DNS-HyXNet achieved up to 99.99% accuracy, with macro-averaged precision, recall, and F1-scores exceeding 99.96%, and demonstrated a per-sample detection latency of just 0.041 ms, confirming its scalability and real-time readiness. These results show that sequential modeling with xLSTM can effectively replace computationally expensive recursive graph generation, offering a deployable and energy-efficient alternative for real-time DNS tunnel detection on commodity hardware.

cs.CV↗

arXiv2025

PULSE: A Unified Multi-Task Architecture for Cardiac Segmentation, Diagnosis, and Few-Shot Cross-Modality Clinical Adaptation

Cardiac image analysis remains fragmented across tasks: anatomical segmentation, disease classification, and grounded clinical report generation are typically handled by separate networks trained under different data regimes. No existing framework unifies these objectives within a single architecture while retaining generalization across imaging modalities and datasets. We introduce PULSE, a multi-task vision-language framework built on self-supervised representations and optimized through a composite supervision strategy that balances region overlap learning, pixel wise classification fidelity, and boundary aware IoU refinement. A multi-scale token reconstruction decoder enables anatomical segmentation, while shared global representations support disease classification and clinically grounded text output allowing the model to transition from pixels to structures and finally clinical reasoning within one architecture. Unlike prior task-specific pipelines, PULSE learns task-invariant cardiac priors, generalizes robustly across datasets, and can be adapted to new imaging modalities with minimal supervision. This moves the field closer to a scalable, foundation style cardiac analysis framework.

cs.CV↗

arXiv2025

Scaling Down to Scale Up: Towards Operationally-Efficient and Deployable Clinical Models via Cross-Modal Low-Rank Adaptation for Medical Vision-Language Models

Foundation models trained via vision-language pretraining have demonstrated strong zero-shot capabilities across diverse image domains, yet their application to volumetric medical imaging remains limited. We introduce MedCT-VLM: Medical CT Vision-Language Model, a parameter-efficient vision-language framework designed to adapt large-scale CT foundation models for downstream clinical tasks. MedCT-VLM uses a parameter-efficient approach to adapt CT-CLIP, a contrastive vision-language model trained on 25,692 chest CT volumes, for multi-label pathology classification using Low-Rank Adaptation (LoRA). Rather than fine-tuning the model's 440 M parameters directly, we insert low-rank decomposition matrices into attention layers of both vision and text encoders, training only 1.67M parameters (0.38\% of total). We evaluate on zero-shot classification across 18 thoracic pathologies, where the model must align CT embeddings with unseen text prompts at inference without task-specific training. LoRA fine-tuning improves mean AUROC from 61.3\% to 68.9\% (+7.6 pp), accuracy from 67.2\% to 73.6\% (+6.4 pp), and macro-F1 from 32.1\% to 36.9\% (+4.8 pp). These results demonstrate that parameter-efficient methods can effectively transfer large-scale pretraining to downstream medical imaging tasks, particularly for zero-shot scenarios where labeled data is scarce.

cs.CV↗

arXiv2025

Generative Adversarial Synthesis and Deep Feature Discrimination of Brain Tumor MRI Images

Compared to traditional methods, Deep Learning (DL) becomes a key technology for computer vision tasks. Synthetic data generation is an interesting use case for DL, especially in the field of medical imaging such as Magnetic Resonance Imaging (MRI). The need for this task since the original MRI data is limited. The generation of realistic medical images is completely difficult and challenging. Generative Adversarial Networks (GANs) are useful for creating synthetic medical images. In this paper, we propose a DL based methodology for creating synthetic MRI data using the Deep Convolutional Generative Adversarial Network (DC-GAN) to address the problem of limited data. We also employ a Convolutional Neural Network (CNN) classifier to classify the brain tumor using synthetic data and real MRI data. CNN is used to evaluate the quality and utility of the synthetic images. The classification result demonstrates comparable performance on real and synthetic images, which validates the effectiveness of GAN-generated images for downstream tasks.

cs.CV↗

arXiv2025

Prompt-Guided Patch UNet-VAE with Adversarial Supervision for Adrenal Gland Segmentation in Computed Tomography Medical Images

Segmentation of small and irregularly shaped abdominal organs, such as the adrenal glands in CT imaging, remains a persistent challenge due to severe class imbalance, poor spatial context, and limited annotated data. In this work, we propose a unified framework that combines variational reconstruction, supervised segmentation, and adversarial patch-based feedback to address these limitations in a principled and scalable manner. Our architecture is built upon a VAE-UNet backbone that jointly reconstructs input patches and generates voxel-level segmentation masks, allowing the model to learn disentangled representations of anatomical structure and appearance. We introduce a patch-based training pipeline that selectively injects synthetic patches generated from the learned latent space, and systematically study the effects of varying synthetic-to-real patch ratios during training. To further enhance output fidelity, the framework incorporates perceptual reconstruction loss using VGG features, as well as a PatchGAN-style discriminator for adversarial supervision over spatial realism. Comprehensive experiments on the BTCV dataset demonstrate that our approach improves segmentation accuracy, particularly in boundary-sensitive regions, while maintaining strong reconstruction quality. Our findings highlight the effectiveness of hybrid generative-discriminative training regimes for small-organ segmentation and provide new insights into balancing realism, diversity, and anatomical consistency in data-scarce scenarios.

eess.IV↗

arXiv2025

Facial Emotion Learning with Text-Guided Multiview Fusion via Vision-Language Model for 3D/4D Facial Expression Recognition

Facial expression recognition (FER) in 3D and 4D domains presents a significant challenge in affective computing due to the complexity of spatial and temporal facial dynamics. Its success is crucial for advancing applications in human behavior understanding, healthcare monitoring, and human-computer interaction. In this work, we propose FACET-VLM, a vision-language framework for 3D/4D FER that integrates multiview facial representation learning with semantic guidance from natural language prompts. FACET-VLM introduces three key components: Cross-View Semantic Aggregation (CVSA) for view-consistent fusion, Multiview Text-Guided Fusion (MTGF) for semantically aligned facial emotions, and a multiview consistency loss to enforce structural coherence across views. Our model achieves state-of-the-art accuracy across multiple benchmarks, including BU-3DFE, Bosphorus, BU-4DFE, and BP4D-Spontaneous. We further extend FACET-VLM to 4D micro-expression recognition (MER) on the 4DME dataset, demonstrating strong performance in capturing subtle, short-lived emotional cues. The extensive experimental results confirm the effectiveness and substantial contributions of each individual component within the framework. Overall, FACET-VLM offers a robust, extensible, and high-performing solution for multimodal FER in both posed and spontaneous settings.

cs.CV↗

arXiv2025

Self-Supervised Multi-View Representation Learning using Vision-Language Model for 3D/4D Facial Expression Recognition

Facial expression recognition (FER) is a fundamental task in affective computing with applications in human-computer interaction, mental health analysis, and behavioral understanding. In this paper, we propose SMILE-VLM, a self-supervised vision-language model for 3D/4D FER that unifies multiview visual representation learning with natural language supervision. SMILE-VLM learns robust, semantically aligned, and view-invariant embeddings by proposing three core components: multiview decorrelation via a Barlow Twins-style loss, vision-language contrastive alignment, and cross-modal redundancy minimization. Our framework achieves the state-of-the-art performance on multiple benchmarks. We further extend SMILE-VLM to the task of 4D micro-expression recognition (MER) to recognize the subtle affective cues. The extensive results demonstrate that SMILE-VLM not only surpasses existing unsupervised methods but also matches or exceeds supervised baselines, offering a scalable and annotation-efficient solution for expressive facial behavior understanding.

cs.CV↗

arXiv2025

Beam-Guided Knowledge Replay for Knowledge-Rich Image Captioning using Vision-Language Model

Generating informative and knowledge-rich image captions remains a challenge for many existing captioning models, which often produce generic descriptions that lack specificity and contextual depth. To address this limitation, we propose KRCapVLM, a knowledge replay-based novel image captioning framework using vision-language model. We incorporate beam search decoding to generate more diverse and coherent captions. We also integrate attention-based modules into the image encoder to enhance feature representation. Finally, we employ training schedulers to improve stability and ensure smoother convergence during training. These proposals accelerate substantial gains in both caption quality and knowledge recognition. Our proposed model demonstrates clear improvements in both the accuracy of knowledge recognition and the overall quality of generated captions. It shows a stronger ability to generalize to previously unseen knowledge concepts, producing more informative and contextually relevant descriptions. These results indicate the effectiveness of our approach in enhancing the model's capacity to generate meaningful, knowledge-grounded captions across a range of scenarios.

cs.CV↗

arXiv2025

Underwater Diffusion Attention Network with Contrastive Language-Image Joint Learning for Underwater Image Enhancement

Underwater images are often affected by complex degradations such as light absorption, scattering, color casts, and artifacts, making enhancement critical for effective object detection, recognition, and scene understanding in aquatic environments. Existing methods, especially diffusion-based approaches, typically rely on synthetic paired datasets due to the scarcity of real underwater references, introducing bias and limiting generalization. Furthermore, fine-tuning these models can degrade learned priors, resulting in unrealistic enhancements due to domain shifts. To address these challenges, we propose UDAN-CLIP, an image-to-image diffusion framework pre-trained on synthetic underwater datasets and enhanced with a customized classifier based on vision-language model, a spatial attention module, and a novel CLIP-Diffusion loss. The classifier preserves natural in-air priors and semantically guides the diffusion process, while the spatial attention module focuses on correcting localized degradations such as haze and low contrast. The proposed CLIP-Diffusion loss further strengthens visual-textual alignment and helps maintain semantic consistency during enhancement. The proposed contributions empower our UDAN-CLIP model to perform more effective underwater image enhancement, producing results that are not only visually compelling but also more realistic and detail-preserving. These improvements are consistently validated through both quantitative metrics and qualitative visual comparisons, demonstrating the model's ability to correct distortions and restore natural appearance in challenging underwater conditions.

cs.CV↗

arXiv2025

Deformable Attentive Visual Enhancement for Referring Segmentation Using Vision-Language Model

Image segmentation is a fundamental task in computer vision, aimed at partitioning an image into semantically meaningful regions. Referring image segmentation extends this task by using natural language expressions to localize specific objects, requiring effective integration of visual and linguistic information. In this work, we propose SegVLM, a vision-language model that incorporates architectural improvements to enhance segmentation accuracy and cross-modal alignment. The model integrates squeeze-and-excitation (SE) blocks for dynamic feature recalibration, deformable convolutions for geometric adaptability, and residual connections for deep feature learning. We also introduce a novel referring-aware fusion (RAF) loss that balances region-level alignment, boundary precision, and class imbalance. Extensive experiments and ablation studies demonstrate that each component contributes to consistent performance improvements. SegVLM also shows strong generalization across diverse datasets and referring expression scenarios.

cs.CV↗