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
- Larry Steven DavisSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works151 from public data
- A Metric Learning Reality Check579
Flaws in the experimental methodology of numerous metric learning papers are found, and it is shown that the actual improvements over time have been marginal at best.
- What Makes Fake Images Detectable? Understanding Properties that Generalize489
This work uses a patch-based classifier with limited receptive fields to visualize which regions of fake images are more easily detectable and shows a technique to exaggerate these detectable properties and demonstrates that, even when the image generator is adversarially finetuned against a fake image classifier, it is still imperfect and leaves detectable artifacts in certain image patches.
- AdaViT: Adaptive Vision Transformers for Efficient Image Recognition355
AdaViT is introduced, an adaptive computation framework that learns to derive usage policies on which patches, self-attention heads and transformer blocks to use throughout the backbone on a per-input basis, aiming to improve inference efficiency of vision transformers with a minimal drop of accuracy for image recognition.
- Making an Invisibility Cloak: Real World Adversarial Attacks on Object Detectors334
A systematic study of adversarial attacks on state-of-the-art object detection frameworks, and a detailed study of physical world attacks using printed posters and wearable clothes, to quantify the performance of such attacks with different metrics.
- MA-LMM: Memory-Augmented Large Multimodal Model for Long-Term Video Understanding329
This study proposes to process videos in an online manner and store past video information in a memory bank, which allows the model to reference historical video content for long-term analysis without exceeding LLMs' context length constraints or GPU memory limits.
- Enhancing Adversarial Example Transferability With an Intermediate Level Attack312
The Intermediate Level Attack (ILA) is introduced, which attempts to fine-tune an existing adversarial example for greater black-box transferability by increasing its perturbation on a pre-specified layer of the source model, improving upon state-of-the-art methods.
- Cross-X Learning for Fine-Grained Visual Categorization277
This paper proposes Cross-X learning, a simple yet effective approach that exploits the relationships between different images and between different network layers for robust multi-scale feature learning and involves two novel components: a cross-category cross-semantic regularizer that guides the extracted features to represent semantic parts and aCross-layerRegularizer that improves the robustness of multi- scale features by matching the prediction distribution across multiple layers.
- 261
- ObjectFormer for Image Manipulation Detection and Localization240
This paper proposes ObjectFormer, a method to detect and localize image manipulations that outperforms state-of-the-art tampering detection and localization methods, and uses a set of learnable object prototypes as mid-level representations to model the object-level consistencies among different regions.
- Detecting Deep-Fake Videos from Appearance and Behavior231
A biometric-based forensic technique for detecting face-swap deep fakes that combines a static biometric based on facial recognition with a temporal, behavioral biometricbased on facial expressions and head movements and is shown to be efficacy across several large-scale video datasets.
- Graph Inductive Biases in Transformers without Message Passing230
The Graph Inductive bias Transformer (GRIT) is proposed -- a new Graph Transformer that incorporates graph inductive biases without using message passing and achieves state-of-the-art empirical performance across a variety of graph datasets, thus showing the power that Graph Transformers without message-passing can deliver.
- On Feature Normalization and Data Augmentation168
This paper proposes Moment Exchange, an implicit data augmentation method that encourages the model to utilize the moment information also for recognition models, and replaces the moments of the learned features of one training image by those of another, and also interpolate the target labels to extract training signal from the moments.
- HNeRV: A Hybrid Neural Representation for Videos166
With content-adaptive embeddings and redesigned architecture, HNeRV outperforms implicit methods in video regression tasks for both reconstruction quality and convergence speed, and shows better internal generalization.
- Generate, Segment, and Refine: Towards Generic Manipulation Segmentation162
A manipulated image generation process that creates true positives using currently available datasets is introduced, and a novel generator for creating examples that force the algorithm to focus on boundary artifacts during training is proposed.
- Open Vocabulary Semantic Segmentation with Patch Aligned Contrastive Learning149
Patch Aligned Contrastive Learning (PACL), a modified compatibility function for CLIP's contrastive loss, is introduced, intending to train an alignment between the patch tokens of the vision encoder and the CLS token of the text encoder, to transfer seamlessly to the task of open vocabulary semantic segmentation.
- On the Robustness of Large Multimodal Models Against Image Adversarial Attacks140
It is found that in general LMMs are not robust to visual adversarial inputs, but the findings suggest that context provided to the model via prompts—such as questions in a QA pair—helps to mitigate the effects of visual adversarial inputs.
- Detecting Everything in the Open World: Towards Universal Object Detection128
UniDetector is proposed, a universal object detector that has the ability to recognize enormous categories in the open world, and surpasses the traditional supervised baselines by more than 4% on average without seeing any corresponding images.
- Quantization Guided JPEG Artifact Correction127
This work creates a novel architecture which is parameterized by the JPEG files quantization matrix, which allows a single model to achieve state-of-the-art performance over models trained for specific quality settings.
- Curriculum Manager for Source Selection in Multi-source Domain Adaptation127
This paper proposed an adversarial agent that learns a dynamic curriculum for source samples, called Curriculum Manager for Source Selection (CMSS), which outperforms other methods on four well-known benchmarks by significant margins.
- 117
- Deep Co-Training with Task Decomposition for Semi-Supervised Domain Adaptation99
It is argued that the labeled target data needs to be distinguished for effective SSDA, and proposed to explicitly decompose the SSDA task into two sub-tasks: a semi-supervised learning (SSL) task in the target domain and an unsupervised domain adaptation (UDA) task across domains.
- 96
- Computationally Budgeted Continual Learning: What Does Matter?95
It is found that, under compute-constrained setting, traditional CL approaches, with no exception, fail to outperform a simple minimal baseline that samples uniformly from memory, suggesting that most existing CL methods are particularly too computationally expensive for realistic budgeted deployment.
- MTFormer: Multi-task Learning via Transformer and Cross-Task Reasoning95
It is demonstrated that models with transformer structures are more appropriate for MTL than convolutional neural networks (CNNs), and a novel transformer-based architecture named MTFormer is proposed, which achieves state-of-the-art results with limited network parameters and computations.
- 89
- Collaborative Real-Time Control of Active Cameras in Large Scale Surveillance Systems83
An algorithm is introduced to address the generic problem of collaboratively controlling a limited number of PTZ cameras to capture an observed number of subjects in an optimal fashion to acquire imagery of subjects for biometrics purposes such as face recognition, or non-facial person identification.
- Unsupervised Domain Adaptation for Semantic Segmentation with GANs.70
This work proposes an approach based on Generative Adversarial Networks (GANs) that brings the embeddings closer in the learned feature space and can achieve state of the art results on two challenging scenarios of synthetic to real domain adaptation.
- Towards Scalable Neural Representation for Diverse Videos69
D-NeRV is proposed, a novel neural representation framework designed to encode diverse videos by decoupling clip-specific visual content from motion information, introducing temporal reasoning into the implicit neural network, and employing the task-oriented flow as intermediate output to reduce spatial redundancies.
- 63
- 53
- Intentonomy: a Dataset and Study towards Human Intent Understanding51
This paper introduces an intent dataset, Intentonomy, comprising 14K images covering a wide range of everyday scenes, and systematically studies whether, and to what extent, commonly used visual information, i.e., object and context, contribute to human motive understanding.
- 48
- Teaching with Soft Label Smoothing for Mitigating Noisy Labels in Facial Expressions45
This work introduces what it calls the Smooth Operator Framework for Teaching (SOFT), based on a mean-teacher architecture where SLS is applied over the teacher’s logits, and finds that the smoothed teacher’s logit provides a beneficial supervision to the student via a consistency loss.
- One-Shot Domain Adaptation for Face Generation45
This paper uses a pre-trained StyleGAN model that already learned the generic face distribution to develop an iterative optimization scheme that rapidly adapts the weights of the model to shift the output's high-level distribution to the target's.
- 45
- Composing Object Relations and Attributes for Image-Text Matching44
Experimental results on two prominent image-text retrieval benchmarks, Flickr30K and MS-COCO, demonstrate that CORA outperforms existing state-of-the-art computationally expensive cross-attention methods regarding recall score while achieving fast computation speed of the dual encoder.
- RegMixup: Mixup as a Regularizer Can Surprisingly Improve Accuracy and Out Distribution Robustness42
The effectiveness of the well celebrated Mixup can be further improved if instead of using it as the sole learning objective, it is utilized as an additional regularizer to the standard cross-entropy loss, which significantly improves the quality of the predictive uncertainty estimation of Mixup in most cases.
- VideoGen-of-Thought: Step-by-step generating multi-shot video with minimal manual intervention41
VideoGen-of-Thought is introduced, a step-by-step framework that automates multi-shot video synthesis from a single sentence by systematically addressing three core challenges: Narrative fragmentation, identity-aware cross-shot propagation, and transition artifacts.
- 41
- Unveiling the Ignorance of MLLMs: Seeing Clearly, Answering Incorrectly34
This investigation identifies two primary issues: 1) most instruction tuning datasets predominantly feature questions that "directly" relate to the visual content, leading to a bias in MLLMs’ responses to other indirect questions, and 2) MLLMs’ attention to visual tokens is notably lower than to system and question tokens.
- 30
- 29
- 27
- 26
- 25
- 25
- 24
- Rapid Adaptation in Online Continual Learning: Are We Evaluating It Right?23
This work proposes a novel metric for measuring adaptation of Online Continual Learning algorithms based on the accuracy on the near-future samples, where spurious correlations are removed, and finds that better generalization can be achieved by retaining and reusing past seen information.
- 22
- DreamDance: Animating Human Images by Enriching 3D Geometry Cues from 2D Poses20
This work presents DreamDance, a novel method for animating human images using only skeleton pose sequences as conditional inputs, and introduces a Mutually Aligned Geometry Diffusion Model to generate fine-grained depth and normal maps for enriched guidance.
- 20
- 19
- 19
- 18
- 17
- 17
- 16
- 16
- 16
- 16
- 15
- 14
- Label Delay in Online Continual Learning13
A new continual learning framework with explicit modeling of the label delay between data and label streams over time steps is introduced, showing experimentally that the method is the least affected by the labeldelay factor and in some cases successfully recovers the accuracy of the non-delayed counterpart.
- 13
- 13
- 12
- 11
- 10
- 10
- 10
- 9
- 9
- 8
- Raising the Bar on the Evaluation of Out-of-Distribution Detection8
This paper defines 2 categories of OoD data using the subtly different concepts of perceptual/visual and semantic similarity to in-distribution (iD) data and proposes a GAN based framework for generating OoD samples from each of these 2 categories, given an iD dataset.
- 8
- 8
- 7
- 7
- 6
- 6
- 6
- Niagara : Normal-Integrated Geometric Affine Field for Scene Reconstruction From a Single View5
The Niagara framework is presented, a new single-view 3D scene reconstruction framework that can faithfully reconstruct challenging outdoor scenes from a single input image for the first time, and substantially improves its ability to capture fine details, mitigating common issues like geometric detail loss and deformation.
- 5
- 5
- 5
- 4
- 4
- 4
- 4
- 4
- From Categories to Classifiers: Name-Only Continual Learning by Exploring the Web3
Overall, this paper underscores the potential of using uncurated webly-supervised data to mitigate the challenges associated with manual data labeling in continual learning.
- 3
- 3
- 3
- 3
- 3
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
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 2026-10-13. 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.