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
- Suggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works207 from public data
- Memorizing Normality to Detect Anomaly: Memory-Augmented Deep Autoencoder for Unsupervised Anomaly Detection1,828
The proposed memory-augmented autoencoder called MemAE is free of assumptions on the data type and thus general to be applied to different tasks and proves the excellent generalization and high effectiveness of the proposed MemAE.
- Graph-Structured Representations for Visual Question Answering496
This paper proposes to build graphs over the scene objects and over the question words, and describes a deep neural network that exploits the structure in these representations, and achieves significant improvements over the state-of-the-art.
- 493
- What Value Do Explicit High Level Concepts Have in Vision to Language Problems?482
A method of incorporating high-level concepts into the successful CNN-RNN approach is proposed, and it is shown that it achieves a significant improvement on the state-of-the-art in both image captioning and visual question answering.
- From Motion Blur to Motion Flow: A Deep Learning Solution for Removing Heterogeneous Motion Blur464
This work directly estimates the motion flow from the blurred image through a fully-convolutional deep neural network (FCN) and recovers the unblurred image from the estimated motion flow and is the first universal end-to-end mapping from the blur image to the dense motion flow.
- Deep learning features at scale for visual place recognition366
This paper trains, at large scale, two CNN architectures for the specific place recognition task and employs a multi-scale feature encoding method to generate condition- and viewpoint-invariant features.
- Adversarial PoseNet: A Structure-Aware Convolutional Network for Human Pose Estimation360
This work proposes a novel structure-aware convolutional network to implicitly take priors about the structure of human bodies into account during training of the deep network and significantly outperforms the state-of-the-art methods and almost always generates plausible human pose predictions.
- Self-supervised Learning of Adversarial Example: Towards Good Generalizations for Deepfake Detection346
This work addresses the generalizable deepfake detection from a simple principle: a generalizable representation should be sensitive to diverse types of forgeries and synthesize augmented forgeries with a pool of forgery configurations and strengthen the “sensitivity” to the forgeries by enforcing the model to predict the forgery configuration.
- ZegCLIP: Towards Adapting CLIP for Zero-shot Semantic Segmentation325
A simpler-and-efficient one-stage solution that directly extends CLIP's zero-shot prediction capability from image to pixel level and demonstrates superior performance, outperforming the state-of-the-art methods by a large margin under both “inductive” and “transductive” zero- shot settings.
- Towards Effective Low-Bitwidth Convolutional Neural Networks266
This paper tackles the problem of training a deep convolutional neural network with both low-precision weights and low-bitwidth activations by proposing a two-stage optimization strategy to progressively find good local minima and adopting a novel learning scheme to jointly train a full- Precision model alongside the low-Precision one.
- METransformer: Radiology Report Generation by Transformer with Multiple Learnable Expert Tokens232
The proposed METransformer, a method to realize a “multi-expert joint diagnosis” mechanism to upgrade the existing “single expert” framework commonly seen in the current literature, enjoys the merits of an ensemble-based approach but through a manner that is computationally more efficient and supports more sophisticated interactions among experts.
- The treasure beneath convolutional layers: Cross-convolutional-layer pooling for image classification201
This paper advocates that if used appropriately, convolutional layer activations constitute a powerful image representation by adopting a new technique proposed in this paper called cross-convolutional-layer pooling, and can achieve comparable or in some cases significantly better performance than existing fully-connected layer based image representations.
- Less is More: Zero-Shot Learning from Online Textual Documents with Noise Suppression194
An l2,1-norm based objective function is proposed which can simultaneously suppress the noisy signal in the text and learn a function to match the text document and visual features and develops an optimization algorithm to efficiently solve the resulting problem.
- Seeing Deeply and Bidirectionally: A Deep Learning Approach for Single Image Reflection Removal191
This work argues that, to remove reflection truly well, it should estimate the reflection and utilize it to estimate the background image, and proposes a cascade deep neural network, which estimates both the Background image and the reflection.
- 189
- Feature Encoding With Autoencoders for Weakly Supervised Anomaly Detection178
This article proposes a novel strategy to transform the input data into a more meaningful representation that could be used for anomaly detection and leverages an autoencoder to encode the data and utilizes three factors, hidden representation, reconstruction residual vector, and reconstruction error, as the new representation for the inputData.
- Towards Context-Aware Interaction Recognition for Visual Relationship Detection168
This paper proposes an alternative, context-aware interaction recognition framework, and explicitly construct an interaction classifier which combines the context, and the interaction, to avoid the limitation of both approaches.
- Structured Binary Neural Networks for Accurate Image Classification and Semantic Segmentation166
This paper proposes a ``network decomposition'' strategy, named Group-Net, in which each full-precision group can be effectively reconstructed by aggregating a set of homogeneous binary branches, and shows strong generalization to other tasks.
- Piecewise Classifier Mappings: Learning Fine-Grained Learners for Novel Categories With Few Examples163
An end-to-end trainable deep network inspired by the state-of-the-art fine-grained recognition model and is tailored for the FSFG task is proposed, which generates the decision boundary via learning a set of more attainable sub-classifiers in a more parameter-economic way.
- Temporal Pyramid Pooling-Based Convolutional Neural Network for Action Recognition157
A novel network structure, which allows an arbitrary number of frames as the network input, is proposed and can be learned on a small target data set because it can leverage the off-the-shelf image-level CNN for model parameter initialization.
- Global and Local Texture Randomization for Synthetic-to-Real Semantic Segmentation128
Two simple yet effective texture randomization mechanisms, GTR and LTR, are proposed to randomize the texture of source images into diverse unreal texture styles to alleviate the reliance of the network on texture while promoting the learning of the domain-invariant cues.
- Semi-supervised Semantic Segmentation with Prototype-based Consistency Regularization122
This work proposes a novel approach to regularize the distribution of within-class features to ease label propagation difficulty and creates a semi-supervised semantic segmentation algorithm that demonstrates superior performance over the state-of-the-art methods from extensive experimental evaluation on both Pascal VOC and Cityscapes benchmarks.
- 119
- Towards using count-level weak supervision for crowd counting114
This paper studies the problem of weakly-supervised crowd counting, and devise a simple-yet-effective training strategy, namely Multiple Auxiliary Tasks Training (MATT), to construct regularizes for restricting the freedom of the generated density maps.
- Mid-level Deep Pattern Mining∗114
This work finds that for an image patch, activation extracted from the first fully-connected layer of a CNN have two appealing properties which enable its seamless integration with pattern mining, and applies this approach to scene and object classification tasks, demonstrating that it outperforms all previous works on mid-level visual element discovery by a sizeable margin.
- 110
- Semi-supervised Crowd Counting via Self-training on Surrogate Tasks93
This paper proposed a novel semi-supervised crowd counting method which is built upon a set of inter-related binary segmentation tasks derived from the original density map regression task as the surrogate prediction target and the surrogate target predictors are learned from both labeled and unlabeled data by utilizing a proposed self-training scheme.
- 93
- Cross-Convolutional-Layer Pooling for Image Recognition93
A novel way to extract image representations from two consecutive convolutional layers: one layer is used for local feature extraction and the other serves as guidance to pool the extracted features, achieving superior performance over existing approaches for extracting image representation from a DCNN.
- Attend in Groups: A Weakly-Supervised Deep Learning Framework for Learning from Web Data91
This work proposes an end-to-end weakly-supervised deep learning framework which is robust to the label noise in Web images and relies on two unified strategies, random grouping and attention, to effectively reduce the negative impact of noisy web image annotations.
- Don't Stop Learning: Towards Continual Learning for the CLIP Model89
A new algorithm is proposed, dubbed Learning without Forgetting via Replayed Vocabulary (VR-LwF), which shows exact effectiveness for alleviating the forgetting issue of the CLIP model.
- Towards Effective Deep Embedding for Zero-Shot Learning88
This paper proposes a pseudo labeling strategy to progressively incorporate the testing samples into the training process and thus balance the model between seen and unseen classes and extends the method to a transductive setting to better handle the model bias problem in ZSL.
- Improved Test-Time Adaptation for Domain Generalization81
This work proposes an Improved Test-Time Adaptation (ITTA) method, and shows that the proposed two strategies are beneficial for the learned model, and ITTA could achieve superior performance to the current state-of-the-art methods on several DG benchmarks.
- Training Quantized Neural Networks With a Full-Precision Auxiliary Module76
The proposed method achieves near lossless performance to the full-precision model by using a 4-bit detector, which is of great practical value, and evaluates the proposed method on image classification and object detection over various quantization approaches and show consistent performance increase.
- 68
- Learning Conditional Attributes for Compositional Zero-Shot Learning63
By encoding conditional attributes, this model enables to generate flexible attribute embeddings for generalization from seen to unseen compositions and demonstrates better performances compared with other state-of-the-art approaches and validate the importance of learning conditional attributes.
- Compositional Model Based Fisher Vector Coding for Image Classification62
This work breaks the convention which assumes that a local feature is drawn from a Gaussian distribution whose mean vector is composed as a linear combination of multiple key components, and the combination weight is a latent random variable and greatly enhance the representative power of the generative model underlying FVC.
- Learning discriminative trajectorylet detector sets for accurate skeleton-based action recognition60
This work shows that the discriminative information of a skeleton sequence usually resides in a short temporal interval and proposes a simple-but-effective local descriptor called trajectorylet to capture the static and kinematic information within this interval and obtains superior results over existing approaches under various experimental setups.
- Image Co-localization by Mimicking a Good Detector’s Confidence Score Distribution59
An entropy-based objective function is devised to enforce the above property when learning the common object detector, and it is shown that despite its simplicity, the approach outperforms state-of-the-arts.
- Don’t Miss the Labels: Label-semantic Augmented Meta-Learner for Few-Shot Text Classification58
This work shows that class-label information can be utilized for extracting more discriminative feature representation of the input text from a pre-trained language model like BERT, and can achieve a performance boost when the samples are scarce.
- Mining Mid-level Visual Patterns with Deep CNN Activations58
It is shown that Convolutional Neural Network activations extracted from image patches typical possess two appealing properties that enable seamless integration with pattern mining techniques, and it is demonstrated that this approach outperforms or matches the performance of the state-of-the-arts on these tasks.
- Effective Training of Convolutional Neural Networks With Low-Bitwidth Weights and Activations55
This paper tackles the problem of training a deep convolutional neural network of both low-bitwidth weights and activations and proposes three practical approaches, including progressive quantization; stochastic precision; and joint knowledge distillation to improve the network training.
- Deep point-to-subspace metric learning for sketch-based 3D shape retrieval52
This work proposes a Deep Point-to-Subspace Metric Learning (DPSML) framework to project a sketch into a feature vector and a 3D shape into a subspace spanned by a few selected basis feature vectors to reduce the redundancy of 3D shapes and introduces a Representative-View Selection (RVS) module to select the most representative views of a3D shape.
- Mask-Aware Networks for Crowd Counting52
This work proposes to use a dedicated network branch to predict the object/non-object mask and then combine its prediction with the input image to produce the density map.
- A Comprehensive Study of GPT-4V’s Multimodal Capabilities in Medical Imaging51
This paper presents a comprehensive evaluation of GPT-4V's capabilities across diverse medical imaging tasks, including Radiology Report Generation, Medical Visual Question Answering (VQA), and Visual Grounding, and finds the limitations of conventional evaluation metrics like the BLEU score.
- Dynamically Transformed Instance Normalization Network for Generalizable Person Re-Identification50
This work proposes a new normalization scheme called Dynamically Transformed Instance Normalization (DTIN) to alleviate the drawback of IN, and employs dynamic convolution to allow the unnormalized feature to control the transformation of the normalized features into new representations.
- Coarse-to-Fine: A RNN-Based Hierarchical Attention Model for Vehicle Re-identification46
This work proposes an end-to-end RNN-based Hierarchical Attention (RNN-HA) classification model for vehicle re-identification and demonstrates that the proposed model achieves superior performance over state-of-the-art methods.
- Progressive Mode-Seeking on Graphs for Sparse Feature Matching45
This paper proposes a simple, yet surprisingly effective approach to explore the huge matching space in order to significantly boost true matches while avoiding outliers and designs a density-aware sampling technique to considerably accelerate mode-seeking.
- Decomposing Disease Descriptions for Enhanced Pathology Detection: A Multi-Aspect Vision-Language Pre-Training Framework42
This paper introduces a novel VLP framework designed to dissect disease descriptions into their fundamental aspects, leveraging prior knowledge about the visual manifestations of pathologies, and presents a dual-head Transformer tailored to process known and unknown diseases, optimizing the comprehensive detection efficacy.
- Adversarial Learning of Structure-Aware Fully Convolutional Networks for Landmark Localization42
This work proposes a novel structure-aware fully convolutional network to implicitly take such priors about the structure of pose components into account during training of the deep network, demonstrating the usefulness of implicit learning of structures using GANs.
- Adaptive Importance Learning for Improving Lightweight Image Super-Resolution Network42
This study presents an adaptive importance learning scheme for SISR that trains the network with an easy-to-complex paradigm by dynamically updating the importance of image pixels on the basis of the training loss, and enables smaller, lightweight, networks to achieve better performance than has previously been possible.
- 40
- 39
- V-PROM: A Benchmark for Visual Reasoning Using Visual Progressive Matrices38
A new large-scale benchmark to evaluates abstract reasoning over real visual data is proposed, and it is found that existing models, including those popular for vision-and-language tasks, are unable to solve seemingly-simple instances.
- Learning Resolution-Adaptive Representations for Cross-Resolution Person Re-Identification37
An alternative SR-free paradigm to directly compare HR and LR images via a dynamic metric that is adaptive to the resolution of a query image is proposed and achieves state-of-the-art performance on multiple CRReID benchmarks.
- Where to Look and How to Describe: Fashion Image Retrieval With an Attentional Heterogeneous Bilinear Network37
The proposed attentional heterogeneous bilinear network (AHBN) consists of two branches: a deep CNN branch to extract fine-grained appearance attributes and a fully convolutional branch to Extract landmark localization information.
- Domain Generalization via Rationale Invariance36
This paper proposes treating the element-wise contributions to the final results as the rationale for making a decision and representing the rationale for each sample as a matrix, indicating the model relies on domain-invariant clues to make decisions, thereby ensuring robust results.
- RPC: a large-scale and fine-grained retail product checkout dataset36
This essay proposes a new dataset, i.e., RPC, which is closer to the realistic setting and can derive a variety of research problems and benchmark the performance on this dataset with a cross-domain detection baseline method.
- 35
- 32
- Learning Discriminative Bayesian Networks from High-Dimensional Continuous Neuroimaging Data31
The contributions of this paper include a new Directed Acyclic Graph (DAG) constraint with theoretical guarantee to ensure the graph validity of GBN, and two general discriminative learning frameworks for Gaussian Bayesian networks (GBN).
- Learning Common Rationale to Improve Self-Supervised Representation for Fine-Grained Visual Recognition Problems29
This paper shows that a common rationale detector can be learned by simply exploiting the GradCAM induced from the SSL objective without using any pre-trained object parts or saliency detectors, making it seamlessly to be integrated with the existing SSL process.
- Sequential Person Recognition in Photo Albums with a Recurrent Network29
This work proposes a novel recurrent network architecture, in which relational information between instances labels and appearance are modeled jointly, and demonstrates that this simple but elegant formulation achieves state-of-the-art performance on the newly released People In Photo Albums (PIPA) dataset.
- Out-of-Distribution Generalization in Natural Language Processing: Past, Present, and Future28
This paper presents the first comprehensive review of recent progress, methods, and evaluations on the generalization challenge from an OOD perspective in natural language understanding, and hopes this survey will encourage future research in this area.
- Weakly Supervised Semantic Segmentation Based on Co-segmentation28
This work proposes a novel method for weakly supervised semantic segmentation with only image-level labels, which relies on a large scale co-segmentation framework that can produce object masks for a group of images containing objects belonging to the same semantic class.
- 27
- Order-aware convolutional pooling for video based action recognition27
Inspired by the capacity of Convolutional Neural Networks in making use of the internal structure of images for information abstraction, this paper proposes to apply the temporal convolution operation to the frame-level representations to extract the dynamic information.
- 27
- 26
- HEp-2 cell image classification with multiple linear descriptors25
This paper achieves 66.6% cell level classification accuracy which is just slightly lower than the best performance achieved in the HEp-2 cell classification contest and proposes a multi-projection-multi-codebook scheme which creates multiple linear projection descriptors and multiple image representation channels with each channel corresponding to one descriptor.
- 25
- CoLeCLIP: Open-Domain Continual Learning via Joint Task Prompt and Vocabulary Learning23
A novel approach is introduced, termed CoLeCLIP, which learns an open-domain CL model based on CLIP, which achieves new state-of-the-art performance for open-domain CL under both task- and class-incremental learning (CIL) settings.
- Structured Binary Neural Networks for Image Recognition23
This paper proposes a “group decomposition” strategy, termed GroupNet, which divides a network into desired groups, and extends the GroupNet for accurate semantic segmentation by embedding the rich context into the binary structure.
- I Learn Better If You Speak My Language: Understanding the Superior Performance of Fine-Tuning Large Language Models with LLM-Generated Responses22
The conclusion reveals that this “familiarity” of an LLM significantly impacts learning performance, and training with LLM-generated responses not only enhances performance but also helps maintain the model’s capabilities in other reasoning tasks after fine-tuning on a specific task.
- 22
- Contextualize Knowledge Bases with Transformer for End-to-end Task-Oriented Dialogue Systems22
This work proposes a COntext-aware Memory Enhanced Transformer framework (COMET), which treats the KB as a sequence and leverages a novel Memory Mask to enforce the entity to only focus on its relevant entities and dialogue history, while avoiding the distraction from the irrelevant entities.
- Incorporation of radius-info can be simple with SimpleMKL21
This paper improves SimpleMKL to utilize the information of the radius of MEB in an efficient and practical way and incorporates its close relative, the trace of data scattering matrix, to avoid the above problems.
- The CLIP Model is Secretly an Image-to-Prompt Converter20
This paper demonstrates that the CLIP model, as utilized in Stable Diffusion, inherently possesses the ability to instantaneously convert images into text prompts, and can be achieved by utilizing a linear projection matrix that is calculated in a closed form.
- 20
- ReFs: A hybrid pre-training paradigm for 3D medical image segmentation18
This paper employs the off-the-shelf medical image segmentation task as reference, and encourages learning a representation that incurs low prediction loss on both SSL and reference tasks and leads to a similar gradient when updating the feature extractor from either task.
- 18
- A Dynamic Parameter Enhanced Network for distant supervised relation extraction18
A novel Dynamic Parameter Enhanced Network (DPEN) for Relation Extraction is developed, which introduces a parameter generator that can dynamically generates the network parameters according to the input query entity types and relation classes.
- Image Captioning with an Intermediate Attributes Layer.18
This work shows that an intermediate image-to-attributes layer can dramatically improve captioning results over the current approach which directly connects an RNN to a CNN.
- 17
- 16
- Discriminative Brain Effective Connectivity Analysis for Alzheimer's Disease: A Kernel Learning Approach upon Sparse Gaussian Bayesian Network16
This paper employs Fisher kernel to bridge the generative models of sparse Bayesian networks (SBN) and the discriminative classifiers of SVMs, and converts the SBN parameter learning to Fisher kernel learning via minimizing a generalization error bound of SVM.
- 15
- Support Vector Machines for Neuroimage Analysis: Interpretation from Discrimination15
This chapter introduces SVM-based methods and applications for neuroimage analysis from this point of view to characterize the group difference captured by SVM with anatomically interpretable patterns, which provides insights into the unknown mechanism of the brain.
- Center Prediction Loss for Re-identification14
A new loss based on center predictivity is proposed, which leads to a more flexible intra-class distribution constraint while ensuring the between-class samples are well-separated and can achieve superior performance and can also be complementary to existing losses.
- 14
- 14
- Effective Tuning Strategies for Generalist Robot Manipulation Policies13
It is observed that in a lowdata regime, with carefully chosen fine-tuning strategies, a GMPs significantly outperforms the state-of-the-art imitation learning algorithms.
- Boundarymix: Generating pseudo-training images for improving segmentation with scribble annotations13
BoundaryMix, a simple-but-effective solution, which generates pseudo training image-annotation pairs from the original images to supplement the missing semantic boundaries and improves the overall segmentation performance.
- 13
- 13
- 13
- A generalized probabilistic framework for compact codebook creation13
The best criterion within the proposed single probabilistic framework achieves the overall best merging performance among the compared merging criteria developed in the literature.
- Zero-Shot Slot Filling with Slot-Prefix Prompting and Attention Relationship Descriptor12
This paper proposes a new prompting scheme that utilizes both learnable tokens and slot names to guide the model to focus on the relevant text spans for a given slot, and uses attention values between tokens to form a feature descriptor for each token.
- 12
- A Causal Inspired Early-Branching Structure for Domain Generalization11
A simple early-branching structure, where the causal and non-causal feature obtaining branches share the first few blocks while diverging thereafter, for better structure design and an effective random domain sampling scheme to fulfill the task.
- Exploring the auxiliary learning for long-tailed visual recognition11
An effective auxiliary learning approach is proposed to split a network into a classifier part and a feature extractor part, and then employ different training strategies for each part in an auxiliary learning manner to promote the awareness of tail-classes.
- 11
- 11
- 11
- S-RRG-Bench: Structured radiology report generation with fine-grained evaluation framework10
This work proposes a specialized evaluation metric (S-Score) that not only measures disease prediction accuracy but also evaluates the precision of disease-specific details, thus offering a clinically meaningful metric for report quality that focuses on elements critical to clinical decision-making and demonstrates a stronger alignment with human assessments.
- MRScore: Evaluating Radiology Report Generation with LLM-based Reward System10
This paper introduces MRScore, an automatic evaluation metric tailored for radiology report generation by leveraging Large Language Models (LLMs), and demonstrates MRScore's higher correlation with human judgments and superior performance in model selection compared to traditional metrics.
- Meta Learning with Differentiable Closed-form Solver for Fast Video Object Segmentation10
This paper tackles the task of fast segmenting all pixels of a target object in all frames, given the annotation mask in the first frame, as a meta-learning problem, where the base learner grasping the semantic scene understanding for a general type of objects, and the meta learner quickly adapting the appearance of the target object with a few examples.
- 10
- 9
- 8
- 8
- 8
- GADE: A Generative Adversarial Approach to Density Estimation and its Applications8
A generative adversarial density estimator (GADE) is proposed, a density estimation approach that bridges the gap between the two likelihood-free models and a Bayesian model where the predictive variance can be leveraged to measure the confidence in the likelihood.
- 8
- 8
- What value high level concepts in vision to language problems8
A method of incorporating high-level concepts into the very successful CNN-RNN approach is proposed, and it is shown that it achieves a significant improvement on the state-of-the-art performance in both image captioning and visual question answering.
- 7
- 7
- A Graph-Embedding Approach to Hierarchical Visual Word Mergence7
This paper proposes a unified hierarchical merging approach built upon the graph-embedding framework that can maintain excellent image classification performance even after a significant dimension reduction, and outperform all the existing comparable visual word-merging methods.
- HiTVideo: Hierarchical Tokenizers for Enhancing Text-to-Video Generation with Autoregressive Large Language Models6
HiTVideo aims to address the potential limitations of existing video tokenizers in text-to-video generation tasks, striving for higher compression ratios, improved token quality, and simplify LLMs modeling under language guidance, offering a scalable and promising framework for advancing text to video generation.
- 6
- 6
- 6
- Pluggable Weakly-Supervised Cross-View Learning for Accurate Vehicle Re-Identification6
This study presents a pluggable Weakly-supervised Cross-View Learning (WCVL) module for vehicle ReID and demonstrates its efficacy by hallucinating the cross-view samples as the hardest positive counterparts with small luminance difference and large local feature variance.
- 6
- A Simple-but-Effective Baseline for Training-Free Class-Agnostic Counting5
This work suggests employing a superpixel algorithm to generate more precise initial point prompts, utilizing an image encoder with richer semantic knowledge to replace the SAM encoder for representing candidate objects, and adopting a multiscale mechanism and a transductive prototype scheme to update the representation of reference examples.
- Enhancing Close-up Novel View Synthesis via Pseudo-labeling5
This approach leverages pseudo-labels derived from existing training data to provide targeted supervision across a wide range of close-up viewpoints and presents a new dataset designed to assess the effectiveness of both current and future methods in this area.
- MuseBarControl: Enhancing Fine-Grained Control in Symbolic Music Generation through Pre-Training and Counterfactual Loss5
A pre-training task designed to link control signals directly with corresponding musical tokens helps in achieving a more effective initialization for subsequent fine-tuning, and a novel counterfactual loss is implemented that promotes better alignment between the generated music and the control prompts.
- Semi-supervised Learning via Conditional Rotation Angle Estimation5
This work proposes a new SemSL approach called Conditional Rotation Angle EStimation (CRAES), which is featured by adopting a module which predicts the image rotation angle conditioned on the candidate image class and achieves superior performance over the other existing ways of combining SlfSL and SemSL.
- 5
- To Balance or Not to Balance: An Embarrassingly Simple Approach for Learning with Long-Tailed Distributions.5
To promote the awareness of tail-classes, a class-balanced sampling scheme is utilised for training both the classifier and the feature extractor, and an auxiliary training task is introduced which is to train a classifier under the regular random sampling scheme.
- 5
- 4
- Close-up-GS: Enhancing Close-Up View Synthesis in 3D Gaussian Splatting with Progressive Self-Training4
This work proposes a novel approach for close-up view generation based by progressively training the 3DGS model with self-generated data and leverages the See3D model, a recently introduced 3D-aware generative model, to enhance the details of rendered views.
- Let Your Video Listen to Your Music! -- Beat-Aligned, Content-Preserving Video Editing with Arbitrary Music4
A novel and efficient framework-termed MVAA (Music-Video Auto-Alignment)-that automatically edits video to align with the rhythm of a given music track while preserving the original visual content while preserving the original visual content is proposed.
- 4
- CAT: A Simple yet Effective Cross-Attention Transformer for One-Shot Object Detection4
The proposed Cross-Attention Transformer (CAT) module utilizes the transformer mechanism to comprehensively capture bi-directional correspondence between any paired pixels from the query and the target image, which empowers the model to sufficiently exploit their semantic characteristics for accurate similarity comparison.
- 4
- Single-Stage Open-world Instance Segmentation with Cross-task Consistency Regularization4
A regularization model to predict foreground pixels and use its relation to instance segmentation to construct a cross-task consistency loss that could alleviate the problem of incomplete instance annotation in Open-World Instance Segmentation.
- Dual Decision Improves Open-Set Panoptic Segmentation4
This paper proposes a divide-and-conquer scheme to develop a two-stage decision process for OPS and shows that by properly combining a known class discriminator with an additional class-agnostic object prediction head, the OPS performance can be improved.
- Generalizable Person Re-Identification via Viewpoint Alignment and Fusion4
This work proposes to use a 3D dense pose estimation model and a texture mapping module to map the pedestrian images to canonical view images and shows that this method can lead to superior performance over the existing approaches in various evaluation settings.
- 4
- 4
- 4
- 4
- Embodied Domain Adaptation for Object Detection3
This work introduces a Source-Free Domain Adaptation (SFDA) approach that adapts a pre-trained model without accessing source data, and refine pseudo labels via temporal clustering, employ multi-scale threshold fusion, and apply a Mean Teacher framework with contrastive learning.
- Unified Unsupervised Anomaly Detection via Matching Cost Filtering3
The UCF framework, a generic post-hoc refinement framework for refining anomaly cost volume of any UAD model, is presented, consistently achieving new state-of-the-art results in both unimodal (RGB) and multimodal (RGB--3D, RGB--Text) UAD scenarios.
- You Can Generate It Again: Data-to-Text Generation with Verification and Correction Prompting3
The potential of using feedback systems to enhance semantic fidelity in smaller language models for data-to-text generation tasks, through the Verification and Correction Prompting (VCP) approach is explored.
- One Last Attention for Your Vision-Language Model3
RAda employs a learned mask, obtained from a lightweight attention layer attached at the end of a VLM, to dynamically calibrate the contribution of each element in the rational matrix, enabling targeted adjustments to the final cross-modal interactions without incurring costly modifications to intermediate features.
- Source-Free Unsupervised Domain Adaptation with Hypothesis Consolidation of Prediction Rationale3
This paper proposes a novel approach that considers multiple prediction hypotheses for each sample and investigates the rationale behind each hypothesis, and identifies the most likely correct hypotheses, which are used as a pseudo-labeled set to support a semi-supervised learning procedure for model adaptation.
- A Generative Approach for Comprehensive Financial Event Extraction at the Document Level3
This paper proposes a generative solution for document-level event extraction, which is more challenging due to the presence of scattered arguments and multiple events, and introduces an encoding scheme to capture entity-to-document level information and a decoding scheme that makes the generative process aware of all relevant contexts.
- 3
- Short-Term Sea Surface Elevation Prediction Using Deep Learning Methods3
This paper examines the performance of Temporal Convolutional Networks (TCNs) and traditional Artificial Neural Networks (ANNs) for short-term ocean wave prediction on the same point wave measurement and illustrates the efficacy of two deep learning models as influenced by nonlinearity and noise levels, but TCNs do not outperform ANNs in theSame point wave prediction.
- Out-of-Distribution Generalization in Text Classification: Past, Present, and Future3
This paper presents the first comprehensive review of recent progress, methods, and evaluations on the generalization challenge from an OOD perspective in text classification, and discusses the challenges involved and potential future research directions.
- 3
- Application of SVMs to the Bag-of-Features Model: A Kernel Perspective3
The fundamentals of the Bag-of-features model in image categorisation are introduced and the novel kernels developed to compare images based on a variety of representations incurred by this model are shown.
- Max-Margin Based Learning for Discriminative Bayesian Network from Neuroimaging Data3
A learning approach to constructing SGBNs that are both representative and discriminative for groups in comparison is proposed, and a max-margin criterion built directly upon the SGBN models is proposed to effectively optimize the classification performance of the S GBNs.
- 2
- 2
- 2
- 2
- RadReason: Radiology Report Evaluation Metric with Reasons and Sub-Scores2
RadReason is introduced, a novel evaluation framework for radiology reports that not only outputs fine-grained sub-scores across six clinically defined error types, but also produces human-readable justifications that explain the rationale behind each score.
- Add-SD: Rational Generation without Manual Reference2
An instruction-based object addition pipeline, named Add-SD, which automatically inserts objects into realistic scenes with rational sizes and positions is introduced, which is solely conditioned on simple text prompts rather than any other human-costly references like bounding boxes.
- Categorical Keypoint Positional Embedding for Robust Animal Re-Identification2
This work introduces an innovative keypoint propagation mechanism, which utilizes a single annotated image and a pre-trained diffusion model to propagate keypoints across an entire dataset, significantly reducing the cost of manual annotation.
- 2
- 2
- Transformer-based Open-world Instance Segmentation with Cross-task Consistency Regularization2
This work proposes a regularization model to predict foreground pixels and use its relation to instance segmentation to construct a cross-task consistency loss that could alleviate the problem of incomplete instance annotation in OWIS datasets.
- Revisiting Image Reconstruction for Semi-supervised Semantic Segmentation2
This paper revisits the idea of using image reconstruction as the auxiliary task and incorporate it with a modern semi-supervised semantic segmentation framework and shows that using reconstruction as auxiliary loss can lead to consistent improvements in various datasets and methods.
- Dilated Convolutional Model for Melody Extraction2
A simple dilated convolutional model that achieves the-state-of-the-art performance on most datasets, for both general and vocal melody extraction, and gets the best performance with the least training data is proposed.
- 2
- 2
- Exploring latent class information for image retrieval using the bag-of-feature model2
A latent class prior is defined in the probabilistic model and derive its marginal distribution, which gives rise to a novel and efficient image similarity measure that can significantly improve retrieval performance without prolonging retrieval process.
- Revisiting Vision–Language Foundations for No-Reference Image Quality Assessment1
This work presents the first systematic evaluation of six prominent pretrained backbones, CLIP, SigLIP2, DINOv2, DINOv3, Perception, and ResNet, for the task of No-Reference Image Quality Assessment (NR-IQA), each fine-tuned using an identical lightweight MLP head.
- 1
- Improving Online Source-Free Domain Adaptation for Object Detection by Unsupervised Data Acquisition1
- PP-SSL : Priority-Perception Self-Supervised Learning for Fine-Grained Recognition1
A novel Priority-Perception Self-Supervised Learning framework, denoted as PP-SSL, which can effectively filter out irrelevant feature interference and extract more subtle discriminative features throughout the training process, and significantly outperforms existing methods across various datasets.
- Pretrained Language Encoders are Natural Tagging Frameworks for Aspect Sentiment Triplet Extraction1
This paper argues that instead of further designing modules to capture the inductive bias of ASTE, PLEs themselves contain "enough" features for 1D and 2D tagging, which can be natural tagging frameworks and achieve a new state of the art, which is verified by extensive experiments and deep analyses.
- Dynamic Memory Enhanced Transformer for End-to-end Task-Oriented Dialogue System1
This work designs a novel Transformer-based Dynamic Memory Network (DMN) with a novel Memory Mask scheme, which can dynamically generate the context-aware knowledge base representations, and reason the knowledge bases simultaneously.
- 1
- A Sketch Based 3D Shape Retrieval Approach Based on Efficient Deep Point-to-Subspace Metric Learning1
- Rethinking Binary Neural Network for Accurate Image Classification and Semantic Segmentation1
This paper proposes to train a network with both binary weights and binary activations, designed specifically for mobile devices with limited computation capacity and power consumption, and claims that considering both value and structure approximation should be the future development direction of BNNs.
- Cross-convolutional-layer Pooling for Generic Visual Recognition.1
A novel way to extract image representations from two consecutive convolutional layers: one layer is utilized for local feature extraction and the other serves as guidance to pool the extracted features.
- 1
- –
- –
- –
- –
- Low-Cost iOS-Based Automated Detection of Under-Construction Interior Drywalls: An Exploratory Study–
- –
- Enhancing Fine-Grained Visual Recognition in the Low-Data Regime Through Feature Magnitude Regularization–
A surprisingly simple regularization technique is introduced to ensure that the magnitudes of the extracted features are evenly distributed, achieved by maximizing the uniformity of feature magnitude distribution, measured through the entropy of the normalized features.
- –
- –
- Progressive Feature Adjustment for Semi-supervised Learning from Pretrained Models–
This paper proposes to use pseudo-labels from the unlabelled data to update the feature extractor that is less sensitive to incorrect labels and only allow the classifier to be trained from the labeled data, and shows that the proposed approach achieves superior performance over existing solutions.
- –
- UOA at the FinNLP-2022 ERAI Task: Leveraging the Class Label Description for Financial Opinion Mining–
This paper summarizes the solutions to the ERAI shared task, which is co-located with the FinNLP workshop at EMNLP 2022, and proposes a BERT-based pre-trained model projecting opinion pairs in a common space for classification.
- –
- –
- –
- DNNRE: A Dynamic Neural Network for Distant Supervised Relation Extraction.–
A novel Dynamic Neural Network for Relation Extraction that can simultaneously handle the style shift problem and leverage mutual class relationships to enhance the prediction accuracy for long-tail classes is developed and can achieve superior performance over the state-of-the-art methods.
- –
- –
- –
- –
- Compact codebook creation and local feature coding in bag-of-features model for image classification–
- –
- HEp-2 Cell Image Classication with Multiple Linear–
A multi-projection-multi-codebook scheme which learns multiple descriptors for representing a same patch and builds multiple BoF models is proposed.
- –
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 2026-10-12. 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.