Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileAcademic lineage
Possible advisorsa guess from early papers, not confirmed
- Amar A. El-SallamSuggested from co-authorship
- Senjian AnSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works121 from public data
- Restormer: Efficient Transformer for High-Resolution Image Restoration4,635
This work proposes an efficient Transformer model by making several key designs in the building blocks (multi-head attention and feed-forward network) such that it can capture long-range pixel interactions, while still remaining applicable to large images.
- Transformers in Vision: A Survey3,900
This survey aims to provide a comprehensive overview of the Transformer models in the computer vision discipline with an introduction to fundamental concepts behind the success of Transformers, i.e., self-attention, large-scale pre-training, and bidirectional feature encoding.
- Multi-Stage Progressive Image Restoration2,345
This paper proposes a novel synergistic design that can optimally balance these competing goals in image restoration tasks, that progressively learns restoration functions for the degraded inputs, thereby breaking down the overall recovery process into more manageable steps.
- Transformers in medical imaging: A survey1,338
This survey surveys the use of Transformers in medical image segmentation, detection, classification, restoration, synthesis, registration, clinical report generation, and other tasks and develops taxonomy for each application.
- Learning Enriched Features for Real Image Restoration and Enhancement1,052
This paper presents a novel architecture with the collective goals of maintaining spatially-precise high-resolution representations through the entire network and receiving strong contextual information from the low- resolution representations, named as MIRNet.
- Cost-Sensitive Learning of Deep Feature Representations From Imbalanced Data990
This paper proposes a cost-sensitive (CoSen) deep neural network, which can automatically learn robust feature representations for both the majority and minority classes, and shows that the proposed approach significantly outperforms the baseline algorithms.
- Learning Enriched Features for Fast Image Restoration and Enhancement653
The approach learns an enriched set of features that combines contextual information from multiple scales, while simultaneously preserving the high-resolution spatial details, and achieves state-of-the-art results for a variety of image processing tasks.
- CycleISP: Real Image Restoration via Improved Data Synthesis460
This paper presents a framework that models camera imaging pipeline in forward and reverse directions that allows any number of realistic image pairs for denoising both in RAW and sRGB spaces and achieves the state-of-the-art performance on real camera benchmark datasets.
- A Self-supervised Approach for Adversarial Robustness385
This paper proposes a self-supervised adversarial training mechanism in the input space that provides significant robustness against the unseen adversarial attacks and can be deployed as a plug-and-play solution to protect a variety of vision systems, as it demonstrates for the case of classification, segmentation and detection.
- 294
- Deep learning based synthesis of MRI, CT and PET: Review and analysis285
This survey comprehensively reviews deep learning-based medical imaging translation from 2018 to 2023 on pseudo-CT, synthetic MR, and synthetic PET, and provides an overview of synthetic contrasts in medical imaging and the most frequently employed deep learning networks for medical image synthesis.
- A Robust Volumetric Transformer for Accurate 3D Tumor Segmentation241
A Transformer architecture for volumetric segmentation, a challenging task that requires keeping a complex balance in encoding local and global spatial cues, and preserving information along all axes of the volume, is proposed.
- Striking the Right Balance With Uncertainty199
This paper demonstrates that the Bayesian uncertainty estimates directly correlate with the rarity of classes and the difficulty level of individual samples, and presents a novel framework for uncertainty based class imbalance learning that efficiently utilizes sample and class uncertainty information to learn robust features and more generalizable classifiers.
- iTAML: An Incremental Task-Agnostic Meta-learning Approach176
A novel meta-learning approach that seeks to maintain an equilibrium between all the encountered tasks, ensured by a new meta-update rule which avoids catastrophic forgetting and is task-agnostic.
- Semantic-Aware Domain Generalized Segmentation174
This paper addresses domain generalized semantic segmentation, where a segmentation model is trained to be domain-invariant without using any target domain data, and proposes a framework including two novel modules: Semantic-Aware Normalization (SAN) and Semantic -Aware Whitening (SAW).
- 172
- Random Path Selection for Continual Learning168
This paper proposes a random path selection algorithm, called RPS-Net, that progressively chooses optimal paths for the new tasks while encouraging parameter sharing and reuse and proposes a simple controller to dynamically balance the model plasticity.
- Adversarial Defense by Restricting the Hidden Space of Deep Neural Networks161
This work proposes to class-wise disentangle the intermediate feature representations of deep networks, and forces the features for each class to lie inside a convex polytope that is maximally separated from the polytopes of other classes.
- AV-Deepfake1M: A Large-Scale LLM-Driven Audio-Visual Deepfake Dataset140
The paper provides a thorough description of the proposed data generation pipeline accompanied by a rigorous analysis of the quality of the generated data and indicates a significant drop in performance compared to previous datasets.
- Regularization of deep neural networks with spectral dropout131
This work introduces a new approach called 'Spectral Dropout' to improve the generalization ability of deep neural networks by eliminating weak and 'noisy' Fourier domain coefficients of the neural network activations, leading to remarkably better results than the current regularization methods.
- An efficient 3D face recognition approach using local geometrical signatures128
This paper presents a computationally efficient 3D face recognition system based on a novel facial signature called Angular Radial Signature (ARS) which is extracted from the semi-rigid region of the face which can handle expression variations.
- Survey: Image mixing and deleting for data augmentation127
An in-depth survey of image mixing and deleting techniques is furnish and essentials, such as each category's strengths and limitations, describing their working mechanism, basic formulations, and applications are presented.
- Self-supervised Knowledge Distillation for Few-shot Learning127
This paper proposes a simple approach to improve the representation capacity of deep neural networks for few-shot learning tasks, and shows that, even in the first stage, self-supervision can outperform current state-of-the-art methods.
- Random Path Selection for Incremental Learning116
This paper proposes a random path selection algorithm, called RPSnet, that progressively chooses optimal paths for the new tasks while encouraging parameter sharing and reuse and proposes a simple controller to dynamically balance the model plasticity.
- 116
- Uncertainty-guided dual-views for semi-supervised volumetric medical image segmentation113
Empirical results suggest that an uncertainty-guided co-training framework can make two neural networks robust to data artefacts and have the ability to generate plausible segmentation masks that can be helpful for semi-automated segmentation processes.
- MARLIN: Masked Autoencoder for facial video Representation LearnINg109
The proposed framework, named MARLIN, is a facial video masked autoencoder that learns highly robust and generic facial embeddings from abundantly available non-annotated web crawled facial videos that can transfer across a variety of facial analysis tasks such as Facial Attribute Recognition (FAR), Facial Expression Recognition, DeepFake Detection, and Lip Synchronization.
- Automatic Gaze Analysis: A Survey of Deep Learning Based Approaches106
This work analyzes recent gaze estimation and segmentation methods, especially in the unsupervised and weakly supervised domain, based on their advantages and reported evaluation metrics and shows that the development of a robust and generic gaze analysis method still needs to address real-world challenges such as unconstrained setup and learning with less supervision.
- 106
- On Generating Transferable Targeted Perturbations104
A new generative approach for highly transferable targeted perturbations (TTP) that not only aligns the global distributions of source and target images, but also matches the local neighbourhood structure between the two domains.
- Orthogonal Projection Loss104
OPL augments the properties of CE loss and directly enforces inter-class separation alongside intra-class clustering in the feature space through orthogonality constraints on the mini-batch level and offers unique advantages e.g., no additional learnable parameters, does not require careful negative mining and is not sensitive to the batch size.
- 99
- A Two-Phase Weighted Collaborative Representation for 3D partial face recognition with single sample97
Experimental results on six challenging 3D facial datasets show that the proposed KMTS-TPWCRC framework achieves promising results for human face recognition with missing parts, occlusions, data corruptions, expressions and pose variations.
- Towards Robust and Reproducible Active Learning using Neural Networks96
It is demonstrated that under identical experimental settings, different types of AL algorithms produce an inconsistent gain over random sampling baseline, and a set of recommendations on how to assess the results using a new AL algorithm to ensure results are reproducible and robust under changes in experimental conditions are concluded.
- Do You Really Mean That? Content Driven Audio-Visual Deepfake Dataset and Multimodal Method for Temporal Forgery Localization96
This work introduces a content-driven audio-visual deepfake dataset, termed Localized Audio Visual DeepFake (LAV-DF), explicitly designed for the task of learning temporal forgery localization and deepfake detection tasks.
- NTIRE 2022 Challenge on Efficient Super-Resolution: Methods and Results94
The NTIRE 2022 challenge was to super-resolve an input image with a magnification factor of ×4 based on pairs of low and corresponding high resolution images and the aim was to design a network for single image super-resolution that achieved improvement of efficiency measured according to several metrics.
- CLIP model is an Efficient Continual Learner88
This work shows that a frozen CLIP (Contrastive Language-Image Pretraining) model offers as-tounding continual learning performance without any zero-shot eval-uation, and advocates the use of this strong yet embarrass-ingly simple baseline for future comparisons in the continual learning tasks.
- 88
- ProposalCLIP: Unsupervised Open-Category Object Proposal Generation via Exploiting CLIP Cues79
The ProposalCLIP is able to predict proposals for a large variety of object categories without annotations, by exploiting CLIP (contrastive language-image pre-training) cues, and shows benefits for downstream tasks, such as unsupervised object detection.
- Synthesizing the Unseen for Zero-Shot Object Detection75
This work proposes a novel generative model that uses class-semantics to not only generate the features but also to discriminatively separate them, using a unified model, to synthesize visual features for unseen classes, so that the model learns both seen and unseen objects in the visual domain.
- A spatial layout and scale invariant feature representation for indoor scene classification74
A new convolutional neural network architecture is designed which incorporates a novel “spatially unstructured” layer to introduce robustness against spatial layout deformations and scale invariance, and a pyramidal image representation is presented.
- Unified Open-Vocabulary Dense Visual Prediction73
A Unified Open-Vocabulary Network (UOVN) is proposed to jointly address four common dense prediction tasks and proposes a multi-modal, multi-scale and multi-task (MMM) decoding mechanism to better exploit multi-modal information for OV recognition.
- 72
- D2-Net: Weakly-Supervised Action Localization via Discriminative Embeddings and Denoised Activations69
A novel loss formulation is introduced, which jointly enhances the discriminability of latent embeddings and robustness of the output temporal class activations with respect to foreground-background noise caused by weak supervision, in a weakly-supervised temporal action localization framework, called D2-Net.
- 63
- Erasing Undesirable Influence in Diffusion Models62
This work introduces EraseDiff, an algorithm designed to preserve the utility of the diffusion model on retained data while removing the unwanted information associated with the data to be forgotten, resulting in a natural first-order algorithm for solving the optimization problem.
- 52
- Investigating Word Affect Features and Fusion of Probabilistic Predictions Incorporating Uncertainty in AVEC 201746
Significant improvements were observed when incorporating prediction uncertainty across various system configurations for predicting arousal and valence, suggesting the importance of taking into consideration prediction uncertainty for fusion and more broadly the advantages of probabilistic predictions.
- 45
- AIM 2020 Challenge on Real Image Super-Resolution: Methods and Results43
This paper introduces the real image Super-Resolution challenge that was part of the Advances in Image Manipulation (AIM) workshop, held in conjunction with ECCV 2020, and gauges the state-of-the-art approaches for real image SR in terms of PSNR and SSIM.
- 41
- Transformer Scale Gate for Semantic Segmentation40
This work proposes a simple yet effective module, Transformer Scale Gate (TSG), to optimally combine multi-scale features in transformer-based segmentation models, and exploits cues in self and cross attentions in Vision Transformers for the scale selection.
- Deformation and Correspondence Aware Unsupervised Synthetic-to-Real Scene Flow Estimation for Point Clouds40
This work develops a point cloud collector and scene flow annotator for GTA-V engine to automatically obtain diverse realistic training samples without human intervention and proposes a mean-teacher-based domain adaptation framework that leverages self-generated pseudo-labels of the target domain.
- Unsupervised Primitive Discovery for Improved 3D Generative Modeling38
A novel factorized generative model for 3D shape generation that sequentially transitions from coarse to fine scale shape generation and an unsupervised primitive discovery algorithm based on a higher-order conditional random field model are introduced.
- 37
- 36
- Protocon: Pseudo-Label Refinement via Online Clustering and Prototypical Consistency for Efficient Semi-Supervised Learning34
Protocon is a novel SSL method aimed at the less-explored label-scarce SSL where such methods usually underperform and delivers significant gains and faster convergence over state-of-the-art across 5 datasets, including CIFARs, ImageNet and DomainNet.
- MTGLS: Multi-Task Gaze Estimation with Limited Supervision34
This work proposes MTGLS: a Multi-Task Gaze estimation framework with Limited Supervision, which leverages abundantly available non-annotated facial image data, and learns strong feature representations of human eyes guided by three complementary auxiliary signals.
- 33
- 30
- Reverse Training: An Efficient Approach for Image Set Classification30
This paper introduces a new approach, called reverse training, to efficiently extend binary classifiers for the task of multi-class image set classification since it trains a single binary classifier to optimally discriminate the class of the query image set from all others.
- Adversarial Training of Variational Auto-Encoders for High Fidelity Image Generation29
This model simultaneously learns to match the data, reconstruction loss and the latent distributions of real and fake images to improve the quality of generated samples and uses an error feedback approach to maintain the equilibrium between competing networks in the model.
- Towards fair decentralized benchmarking of healthcare AI algorithms with the Federated Tumor Segmentation (FeTS) challenge28
The authors describe the Federated Tumor Segmentation (FeTS) challenge for the decentralised benchmarking of FL algorithms and evaluation of Healthcare AI algorithm generalizability in real-world cancer imaging datasets.
- Learned 3D Shape Representations Using Fused Geometrically Augmented Images: Application to Facial Expression and Action Unit Detection27
The proposed approach to learn generic multi-modal mesh surface representations using a novel scheme for fusing texture and geometric data is both computationally and memory efficient, preserves intrinsic geometric information and learns highly discriminative feature representations by effectively fusing shape and texture information at data level.
- Open-Vocabulary Object Detection via Scene Graph Discovery25
This paper proposes a novel Scene-Graph-Based Discovery Network (SGDN) that exploits scene graph cues for OV detection and proposes scene-graph-based prediction (SGPred), where it takes scene graphs as bridges to improve the consistency between cross-modal embeddings for Ov object classification.
- Empowering Simple Binary Classifiers for Image Set Based Face Recognition25
This paper suggests a different approach which does not resort to a single representation of an image set and an efficient classification strategy is developed which extends well-known simple binary classifiers for the task of multi-class image set classification.
- OCAI: Improving Optical Flow Estimation by Occlusion and Consistency Aware Interpolation24
This work proposes OCAI, a method that supports robust frame interpolation by generating intermediate video frames along-side optical flows in between and introduces a teacher-student style semi-supervised learning method on top of the interpolated frames.
- 24
- Stylized Adversarial Defense21
This work uses the style and content information of the target sample from another class, alongside its class-boundary information, to create adversarial perturbations, and proposes a max-margin adversarial training approach that minimizes the distance between source image and its adversary and maximizes thedistance between the adversary and the target image.
- 20
- Concealing Sensitive Samples against Gradient Leakage in Federated Learning18
This work proposes synthesizing concealed samples to mimic the sensitive data at the gradient level while ensuring their visual dissimilarity from the actual sensitive data while simultaneously maintaining the FL performance.
- 16
- 16
- 16
- CustomKD: Customizing Large Vision Foundation for Edge Model Improvement via Knowledge Distillation15
Beyond providing well-generalized original knowledge from teachers, CustomKD aligns the features of teachers to those of students, making it easy for students to understand and overcome the large model discrepancy overall.
- 15
- Unpaired referring expression grounding via bidirectional cross-modal matching14
A novel bidirectional cross-modal matching (BiCM) framework that introduces top-down perspective via generating query-specific visual attention maps and a knowledge adaptation matching (KAM) module that leverages unpaired training data to adapt pretrained knowledge to the target dataset and task.
- 13
- 13
- 1M-Deepfakes Detection Challenge12
The 1M-Deepfakes Detection Challenge is designed to engage the research community in developing advanced methods for detecting and localizing deepfake manipulations within the large-scale high-realistic audio-visual dataset.
- 12
- 12
- 12
- DuoLoRA: Cycle-Consistent and Rank-Disentangled Content-Style Personalization11
This work proposes DuoLoRA-a content-style personalization framework featuring three key components: rank-dimension mask learning, effective merging via layer priors, and Constyle loss, which leverages cycle-consistency in the merging process.
- PosSAM: Panoptic Open-vocabulary Segment Anything11
An open-vocabulary panoptic segmentation model that effectively unifies the strengths of the Segment Anything Model with the vision-language CLIP model in an end-to-end framework and proposes a novel Local Discriminative Pooling (LDP) module leveraging class-agnostic SAM and class-aware CLIP features for unbiased open-vocabulary classification.
- Real-Time Trajectory-Based Social Group Detection10
This approach explores the impact of motion trajectory on social grouping and utilizes a novel, reliable, and fast data-driven method that is significantly faster, with up to 12x faster inference times compared to state-of-the-art methods under the same computation resources.
- Meta-learning the Learning Trends Shared Across Tasks10
This work proposes a 'Path-aware' model-agnostic meta-learning approach that not only learns a transferable initialization for adaptation, but also models the optimal update directions, learning rates, and task-specific learning trends.
- Conditional Distribution Modelling for Few-Shot Image Synthesis with Diffusion Models7
Conditional Distribution Modelling (CDM) is proposed -- a framework which effectively utilizes Diffusion models for few-shot image generation, and a novel inversion based optimization strategy is devised that improves the approximated unseen class distribution, and ensures the fidelity of the generated samples to the unseen class.
- A multimodal system to characterise melancholia: cascaded bag of words approach7
This study is the first attempt to propose a multimodal system to differentiate melancholia from non-melancholia and healthy controls and reports the sensitivity and specificity of classification in depressive subtypes.
- Flashbacks to harmonize stability and plasticity in continual learning6
Flashback Learning is introduced, a novel method designed to harmonize the stability and plasticity of models in Continual Learning (CL), which outperforms state-of-the-art CL methods on more challenging datasets like ImageNet.
- FLoC: Facility Location-Based Efficient Visual Token Compression for Long Video Understanding6
FLoC is an efficient visual token compression framework based on the facility location function, a principled approach that swiftly selects a compact yet highly representative and diverse subset of visual tokens within a predefined budget on the number of visual tokens.
- 6
- HyperNet Fields: Efficiently Training Hypernetworks without Ground Truth by Learning Weight Trajectories5
This work proposes a method to train hypernetworks, without the need for any per-sample ground truth, and introduces an additional input to the Hypernetwork, the convergence state, which makes it act as a neural field that models the entire convergence pathway of a task network.
- GraphITTI: Attributed Graph-based Dominance Ranking in Social Interaction Videos5
This paper aims to reformulate the problem of detecting the Most Dominant Person (MDP) as a person ranking problem by utilizing person-specific attributes such as facial gestures, eye gaze, visual attention and speaking patterns, to inject domain knowledge into the GraphITTI framework.
- 5
- 5
- Tripartite Weight-Space Ensemble for Few-Shot Class-Incremental Learning4
This work introduces a novel FSCIL method to effectively address catastrophic forgetting and overfitting issues and proposes a tripartite weight-space ensemble (Tri-THE AUTHORS), which enables to seamlessly update the entire model with a few examples.
- SubZero: Composing Subject, Style, and Action via Zero-Shot Personalization4
This work presents SubZero, a novel framework to generate any subject in any style, performing any action without the need for fine-tuning, and proposes a novel set of constraints to enhance subject and style similarity, while reducing leakage.
- Emolysis: A Multimodal Open-Source Group Emotion Analysis and Visualization Toolkit4
Emolysis is a Python-based, standalone open-source group emotion analysis toolkit for use in different social situations upon getting consent from the users and encourages application developers to extend it to application specific environments on top of the existing system.
- AV-GAZE: A Study on the Effectiveness of Audio Guided Visual Attention Estimation for Non-profilic Faces4
If audio-guided coarse head-pose can further enhance visual attention estimation performance for non-prolific faces and, when AV-Gaze is tested on benchmark datasets with these specific modalities, it achieves competitive results on multiple datasets, while being highly adaptive toward challenging scenarios.
- ConsNoTrainLoRA: Data-driven Weight Initialization of Low-Rank Adapters Using Constraints3
This paper improves convergence and final performance of LoRA finetuning, using the proposed data-driven weight initialization method, ConsNoTrainLoRA (CNTLoRA), and demonstrates that CNTLoRA outperforms standard and data-driven weight initialization methods.
- 3
- 3
- 3
- 2
- 2
- LAVA:Label-efficient Visual Learning and Adaptation2
LAVA learns self-supervised visual representations on the source dataset and ground them using class label semantics to overcome transfer collapse problems associated with supervised pretraining and maximises the gains from unlabelled target data via a novel method which uses multi-crop augmentations to obtain highly robust pseudo-labels.
- Memory-Efficient Fine-Tuning Diffusion Transformers via Dynamic Patch Sampling and Block Skipping1
A memory-efficient fine-tuning framework called DiT-BlockSkip, integrating timestep-aware dynamic patch sampling and block skipping by precomputing residual features, which achieves competitive personalization performance qualitatively and quantitatively, while reducing memory usage substantially.
- Personalized OVSS: Understanding Personal Concept in Open-Vocabulary Semantic Segmentation1
This paper introduces a novel task termed personalized open-vocabulary semantic segmentation and proposes a text prompt tuning-based plug-in method designed to recognize personal visual concepts using a few pairs of images and masks, while maintaining the performance of the original OVSS.
- 1
- 1
- 1
- –
- –
- –
- DiffAugment: Diffusion Based Long-Tailed Visual Relationship Recognition–
DiffAugment is introduced -- a method which first augments the tail classes in the linguistic space by making use of WordNet and then utilizes the generative prowess of Diffusion Models to expand the visual space for minority classes.
- –
- –
- –
- Learning techniques for multi-modal facial analysis–
This dissertation proposes an automatic framework, which exploits the dynamics of textured 3D videos for the recognizing six discrete facial expressions, and proposes manifold learning, deep learning and discriminative learning techniques for face recognition across multiple data modalities.
- –
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-15. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.