Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks164 from public data
- 8,503
- 1,613
- 920
- 391
- WebFace260M: A Benchmark Unveiling the Power of Million-Scale Deep Face Recognition375
A distributed framework is developed to train face recognition models efficiently without tampering with the performance, and comprehensive baselines are established on a rich-attribute test set under FRUITS-100ms/500ms/1000ms protocol.
- Sub-center ArcFace: Boosting Face Recognition by Large-Scale Noisy Web Faces305
The intra-class constraint of ArcFace is relaxed to improve the robustness to label noise and recapture intra-compactness, decrease the influence from noise, and achieve comparable performance compared to ArcFace trained on the manually cleaned dataset.
- Dual T: Reducing Estimation Error for Transition Matrix in Label-noise Learning298
This paper introduces an intermediate class to avoid directly estimating the noisy class posterior of the transition matrix, and introduces the dual $T-estimator for estimating transition matrices, leading to better classification performances.
- 224
- UV-GAN: Adversarial Facial UV Map Completion for Pose-Invariant Face Recognition215
A framework for training Deep Convolutional Neural Network (DCNN) to complete the facial UV map extracted from in-the-wild images, and devise a meticulously designed architecture that combines local and global adversarial DCNNs to learn an identity-preserving facial UV completion model.
- WiLoR: End-to-end 3D Hand Localization and Reconstruction in-the-wild204
This work presents a data-driven pipeline for efficient multi-hand reconstruction in the wild, composed of a real-time fully convolutional hand localization and a high-fidelity transformer-based 3D hand reconstruction model.
- 169
- Spatio-temporal convolutional features with nested LSTM for facial expression recognition161
A novel end-to-end architecture termed Spatio-Temporal Convolutional features with Nested LSTM (STC-NLSTM), which learns the muti-level appearance features and temporal dynamics of facial expressions in a joint fashion is proposed.
- 149
- Deep Polynomial Neural Networks142
Three tensor decompositions that significantly reduce the number of parameters and show how they can be efficiently implemented by hierarchical neural networks are introduced.
- 135
- Killing Two Birds with One Stone: Efficient and Robust Training of Face Recognition CNNs by Partial FC128
A sparsely updating variant of the FC layer, named Partial FC (PFC), where all class centers are still maintained throughout the whole training process, but only a subset is selected and updated in each iteration, resulting in a dramatically reduced computing requirement and frequency of passive update on tail class centers.
- Masked Face Recognition Challenge: The InsightFace Track Report120
This workshop organizes Masked Face Recognition (MFR) challenge1 and focuses on bench-marking deep face recognition methods under the existence of facial masks, and builds up an online model testing system, which can give a comprehensive evaluation of face recognition models.
- Joint Multi-View Face Alignment in the Wild103
This paper proposes the first, to the best of the knowledge, joint multi-view convolutional network to handle large pose variations across faces in-the-wild, and elegantly bridge face detection and facial landmark localization tasks.
- The Menpo Benchmark for Multi-pose 2D and 3D Facial Landmark Localisation and Tracking101
An elaborate semi-automatic methodology is introduced for providing high-quality annotations for both the Menpo 2D and Menpo 3D benchmarks, two new datasets for multi-pose 2d and 3D facial landmark localisation and tracking.
- 100
- ALIP: Adaptive Language-Image Pre-training with Synthetic Caption99
An Adaptive Language-Image Pre-training (ALIP), a bi-path model that integrates supervision from both raw text and synthetic caption that achieves state-of-the-art performance on multiple downstream tasks including zero-shot image-text retrieval and linear probe is proposed.
- CFPL-FAS: Class Free Prompt Learning for Generalizable Face Anti-Spoofing93
This work proposes a novel Class Free Prompt Learning (CFPL) paradigm for DG FAS, which utilizes two lightweight transformers, namely Content Q-Former (CQF) and Style Q-Former (SQF), to learn the different semantic prompts conditioned on content and style features by using a set of learnable query vectors.
- Breaking the Modality Barrier: Universal Embedding Learning with Multimodal LLMs88
UniME (Universal Multimodal Embedding), a novel two-stage framework that leverages MLLMs to learn discriminative representations for diverse downstream tasks, and introduces hard negative enhanced instruction tuning to further advance discriminative representation learning.
- Boosting Object Detection with Zero-Shot Day-Night Domain Adaptation85
This work proposes to boost low-light object detection with zero-shot day-night domain adaptation, which aims to generalize a detector from well-lit scenarios to low-light ones without requiring real low-light data.
- Redesigning Multi-Scale Neural Network for Crowd Counting84
A new relative local counting loss based on relative count differences among hard-predicted local regions in an image, which proves to be complementary to the conventional absolute error loss on the density map, is introduced.
- Neural Sign Actors: A diffusion model for 3D sign language production from text82
This work can generate dynamic sequences of 3D avatars from an un-constrained domain of discourse using a diffusion process formed on a novel and anatomically informed graph neu-ral network defined on the SMPL-X body skeleton.
- HaWoR: World-Space Hand Motion Reconstruction from Egocentric Videos81
This work proposes HaWoR, a high-fidelity method for hand motion reconstruction in world coordinates from egocentric videos, and proposes an adaptive egocentric SLAM framework that addresses the shortcomings of traditional SLAM methods, providing robust performance under challenging camera dynamics.
- Dense 3D Face Decoding Over 2500FPS: Joint Texture & Shape Convolutional Mesh Decoders81
This paper presents the first, to the best of the knowledge, non-linear 3DMMs by learning joint texture and shape auto-encoders using direct mesh convolutions and demonstrates how these auto- encoders can be used to train very light-weight models that perform Coloured Mesh Decoding in-the-wild at a speed of over 2500 FPS.
- Robotic Perception of Transparent Objects: A Review80
This article provides a comprehensive survey of the platforms and recent advances for robotic perception of transparent objects, and highlights the main challenges and proposed future directions of various transparent object perception tasks, i.e., segmentation, reconstruction, and pose estimation.
- Understanding and Mitigating Overfitting in Prompt Tuning for Vision-Language Models71
Subspace Prompt Tuning (Sub PT) is proposed to project the gradients in back-propagation onto the low-rank subspace spanned by the early-stage gradient flow eigenvectors during the entire training process and successfully eliminate the overfitting problem.
- $\Pi-$nets: Deep Polynomial Neural Networks71
It is empirically demonstrated that $\Pi$-Nets have better representation power than standard DCNNs and they even produce good results without the use of non-linear activation functions in a large battery of tasks and signals, i.e., images, graphs, and audio.
- P–nets: Deep Polynomial Neural Networks71
It is empirically demonstrated that $\Pi$-Nets have better representation power than standard DCNNs and they even produce good results without the use of non-linear activation functions in a large battery of tasks and signals, i.e., images, graphs, and audio.
- 69
- 63
- OSTeC: One-Shot Texture Completion62
This work proposes an unsupervised approach for one-shot 3D facial texture completion that does not re-quire large-scale texture datasets, but rather harnesses the knowledge stored in 2D face generators to improve pose-invariant face recognition.
- 61
- Unicom: Universal and Compact Representation Learning for Image Retrieval60
This paper clusters the large-scale LAION400M into one million pseudo classes based on the joint textual and visual features extracted by the CLIP model and randomly selects partial inter-class prototypes to construct the margin-based softmax loss, making the classification conflict-robust and the feature embedding compact.
- 60
- M3 CSR: Multi-view, multi-scale and multi-component cascade shape regression58
A multi-view, multi-scale and multi-component cascade shape regression (M3CSR) model for robust face alignment is presented and a component-based shape refinement process is developed to further improve the performance of face alignment.
- 56
- 56
- Facial Geometric Detail Recovery via Implicit Representation55
This work presents a robust texture-guided geometric detail recovery approach using only a single in-the-wild facial image and registers the implicit shape details to a 3D Morphable Model template, which can be used in traditional modeling and rendering pipelines.
- FitMe: Deep Photorealistic 3D Morphable Model Avatars53
FitMe achieves state-of-the-art reflectance acquisition and identity preservation on single “in- the-wild” facial images, while it produces impressive scan-like results, when given multiple unconstrained facial images pertaining to the same identity.
- Decoupled Multi-task Learning with Cyclical Self-Regulation for Face Parsing52
A novel Decoupled Multi-task Learning with Cyclical Self-Regulation (DML-CSR) for face parsing which comprises face parsing, binary edge, and category edge detection and develops a dynamic dual graph convolutional network to capture global contextual information without using any extra pooling operation.
- $V_{k}D$: Improving Knowledge Distillation Using Orthogonal Projections47
A novel constrained feature distillation method derived from a small set of core principles, which results in two emerging components: an orthogonal projection and a task-specific normalisation that can outperform all previous methods on ImageNet and reach up to a 4.4% relative improvement over the previous state-of-the-art methods.
- 47
- 45
- 44
- IDAdapter: Learning Mixed Features for Tuning-Free Personalization of Text-to-Image Models42
The introduced IDAdapter is a tuning-free approach that enhances the diversity and identity preservation in personalized image generation from a single face image, achieving both diversity and identity fidelity.
- 38
- TransFace++: Rethinking the Face Recognition Paradigm With a Focus on Accuracy, Efficiency, and Security37
This work proposes a superior FR model called TransFace, which contains a patch-level data augmentation strategy named Dominant Patch Amplitude Perturbation (DPAP) and a hard sample mining strategy named Entropy-guided Hard Sample Mining (EHSM), and investigates the intrinsic property of image bytes, to improve inference efficiency and user privacy protection.
- DiffTAD: Temporal Action Detection with Proposal Denoising Diffusion37
A new formulation of temporal action detection with denoising diffusion with denoising diffusion, DiffTAD, which can yield action proposals accurately given an untrimmed long video, presents a generative modeling perspective, against previous discriminative learning manners.
- Augmenting Deep Classifiers with Polynomial Neural Networks34
This work cast the study of deep classifiers under a unifying framework and expresses state-of-the-art architectures in the form of different degree polynomials of the input in order to provide insights on the inductive biases of each model and enables natural extensions building upon their polynomial nature.
- 34
- WaveFace: Authentic Face Restoration with Efficient Frequency Recovery29
WaveFace is proposed to solve the problems in the frequency domain, where low- and high-frequency components decomposed by wavelet transformation are considered individually to maximize authenticity as well as efficiency.
- Wild Face Anti-Spoofing Challenge 2023: Benchmark and Results29
The Wild Face Anti-Spoofing dataset is introduced, a large-scale, diverse FAS dataset collected in unconstrained settings that markedly enhances FAS data diversity and hosts the WFAS Challenge at the CVPR2023 workshop.
- MogFace: Towards a Deeper Appreciation on Face Detection29
In this paper, three key components, Adaptive Online Incremental Anchor Mining Strategy, Selective Scale Enhancement Strategy and Hierarchical Context-Aware Module are separately proposed to boost the performance of face detectors.
- Transferring Annotator- and Instance-Dependent Transition Matrix for Learning From Crowds28
Theoretical analyses are derived to demonstrate that both the knowledge transfer from global to individuals and the knowledge transfer between neighboring individuals can effectively help mitigate the challenge of modeling general AIDTM.
- Extended : Learning with Mixed Closed-set and Open-set Noisy Labels28
The traditional transition matrix is extended to be able to model mixed label noise, and further to the cluster-dependent extended transition matrix to better combat the instance-dependent label noise in real-world applications.
- Side Information for Face Completion: A Robust PCA Approach28
This paper proposes two models for the RPCA that take into account side information, even in the presence of missing values, and constructs a generative adversarial network (GAN) to extract side information as well as subspaces.
- “Principal Components” Enable a New Language of Images27
A novel visual tokenization framework that embeds a provable PCA-like structure into the latent token space and identifies and resolved a semantic-spectrum coupling effect that causes the unwanted entanglement of high-level semantic content and low-level spectral details in the tokens by leveraging a diffusion decoder.
- Gait Recognition in the Wild: A Large-Scale Benchmark and NAS-Based Baseline27
The proposed GREW benchmark proves to be essential for both training and evaluating gait recognizers in unconstrained scenarios, and the Single Path One-Shot neural architecture search with uniform sampling for Gait recognition is proposed, named SPOSGait, which is the first NAS-based gait recognition model.
- 3DGazeNet: Generalizing 3D Gaze Estimation with Weak-Supervision from Synthetic Views27
This work proposes to train general gaze estimation models which can be directly employed in novel environments without adaptation, and leverage the observation that head, body, and hand pose estimation benefit from revising them as dense 3D coordinate prediction, and similarly express gaze estimation as regression of dense 3D eye meshes.
- ImagiDrive: A Unified Imagination-and-Planning Framework for Autonomous Driving26
ImagiDrive is a novel end-to-end autonomous driving framework that integrates a VLM-based driving agent with a DWM-based scene imaginer to form a unified imagination-and-planning loop and introduces an early stopping mechanism and a trajectory selection strategy to address efficiency and predictive accuracy challenges inherent in this integration.
- 26
- Robust facial landmark tracking via cascade regression26
A robust facial landmark tracking system via cascade regression, which is able to deal well with some challenges emerging in the sequential images, and achieves competing results on the LFPW, Helen, 300-W and 300-VW datasets.
- EDFace-Celeb-1 M: Benchmarking Face Hallucination With a Million-Scale Dataset25
A public Ethnically Diverse Face dataset is built, EDFace-Celeb-1 M, and a benchmark task for face hallucination is designed, demonstrating the performance and limitations of state-of-the-art algorithms.
- 25
- 24
- Regularization of polynomial networks for image recognition23
This work introduces a class of PNs, which are able to reach the performance of ResNet across a range of six benchmarks and introduces D-PolyNets that achieve a higher- degree of expansion than previously proposed polynomial networks.
- 23
- CEDex: Cross-Embodiment Dexterous Grasp Generation at Scale from Human-like Contact Representations22
CEDex is proposed, a novel cross-embodiment dexterous grasp synthesis method at scale that bridges human grasping kinematics and robot kinematics by aligning robot kinematic models with generated human-like contact representations.
- Inverse Image Frequency for Long-Tailed Image Recognition21
Inverse Image Frequency is a multiplicative margin adjustment transformation of the logits in the classification layer of a convolutional neural network that achieves stronger performance than similar works and it is especially useful for downstream tasks such as long-tailed instance segmentation as it produces fewer false positive detections.
- 21
- 21
- 21
- 20
- Multi-label Cluster Discrimination for Visual Representation Learning19
A novel Multi-Label Cluster Discrimination method named MLCD to enhance representation learning and achieves state-of-the-art performance on multiple downstream tasks including linear probe, zero-shot classification, and image-text retrieval.
- Self-Adaptive Reality-Guided Diffusion for Artifact-Free Super-Resolution19
This work proposes Self-Adaptive Reality-Guided Diffusion (SARGD), a training-free method that delves into the latent space to effectively identify and mitigate the propagation of artifacts, and introduces a Self-Adaptive Guidance (SAG) mechanism that dynamically computes a reality score, enhancing the sharpness of the realistic latent.
- Region-based Cluster Discrimination for Visual Representation Learning18
This work introduces Region-Aware Cluster Discrimination (RICE), a novel method that enhances region-level visual and OCR capabilities and designs a unified region cluster discrimination loss that jointly supports object and OCR learning within a single classification framework, enabling efficient and scalable distributed training on large-scale data.
- 18
- 18
- 18
- 17
- Training tactile sensors to learn force sensing from each other16
GenForce is presented, the first framework that enables transferable force sensing across diverse tactile sensors in robotic hands using a unified representation, enhancing robot manipulation through cross-sensor transfer and multi-sensor coordination.
- 16
- Deep Face Restoration: A Survey16
This paper comprehensively surveys recent advances in deep learning techniques for face restoration and presents a comprehensive review of existing FR methods, including prior based methods and deep learning-based methods.
- MimicME: A Large Scale Diverse 4D Database for Facial Expression Analysis16
MimicMe, a novel large-scale database of dynamic high-resolution 3D faces with great diversity on age, gender and ethnicity, is introduced and it is envisioned that it will be extremely valuable to researchers working in many problems of face modelling and analysis.
- Cascaded Regional Spatio-Temporal Feature-Routing Networks for Video Object Detection16
This paper incorporates the correlation filter tracking on the convolutional feature maps to explore an efficient and effective spatiotemporal region proposal generation method, and proposes a router function which selects feature maps according to the scale of proposals.
- Spatio-temporal Prompting Network for Robust Video Feature Extraction15
A neat and unified framework, called Spatio-Temporal Prompting Network (STPN), that can efficiently extract robust and accurate video features by dynamically adjusting the input features in the backbone network and is easy to generalise to various video tasks because it does not contain task-specific modules.
- Poly-NL: Linear Complexity Non-local Layers With 3rd Order Polynomials15
This work overcome the efficiency limitation of non-local blocks by framing them as special cases of 3rd order polynomial functions, which enables them to formulate novel fast Non-Local blocks capable of reducing the complexity from quadratic to linear with no loss in performance.
- Confidence-Guided Centroids for Unsupervised Person Re-Identification14
Confidence-Guided Centroids (CGC) are proposed to provide reliable cluster-wise prototypes for feature learning and Confidence-Guided pseudo Label (CGL), which enables samples to approach not only the originally assigned centroid but also other centroids that are potentially embedded with their identity information.
- From Attention to Activation: Unravelling the Enigmas of Large Language Models14
These methods reduce the attention proportion on the first token from 65% to 3.3%, the activation kurtosis in the hidden states from 1657 to 3.1, and perplexity penalty under 4-bit weight quantisation from 3565 to 0.3.
- FaceHunter: A multi-task convolutional neural network based face detector14
The proposed FaceHunter is evaluated on the AFW dataset, FDDB dataset and Pascal Faces respectively, and extensive experiments demonstrate its powerful performance against several state-of-the-art detectors.
- ZeroGS: Training 3D Gaussian Splatting from Unposed Images13
This work proposes ZeroGS to train 3DGS from hundreds of unposed and unordered images, and recovers more accurate camera poses than state-of-the-art pose-free NeRF/3DGS methods, and even renders higher quality images than 3DGS with COLMAP poses.
- Adaptive Parametric Activation13
The Adaptive Parametric Activation (APA) function is proposed, a novel and versatile activation function that unifies most common activation functions under a single formula that increases the performance in multiple benchmarks across various model architectures.
- 12
- FaceMAE: Privacy-Preserving Face Recognition via Masked Autoencoders12
A novel framework FaceMAE, where the face privacy and recognition performance are considered simultaneously, is proposed, which consistently reduces at least 50% error rate on LFW, CFP-FP and AgeDB.
- Embodied Arena: A Comprehensive, Unified, and Evolving Evaluation Platform for Embodied AI11
Embodied Arena is presented, a comprehensive, unified, and evolving evaluation platform for Embodied AI that establishes a systematic embodied capability taxonomy spanning three levels (perception, reasoning, task execution), seven core capabilities, and 25 fine-grained dimensions, enabling unified evaluation with systematic research objectives.
- Fractal Calibration for long-tailed object detection11
FRActal CALibration (FRACAL) is a novel post-calibration method for long-tailed object detection that boosts the rare class performance by up to 8.6% and surpasses all previous methods on LVIS dataset, while showing good generalisation to other datasets such as COCO, V3Det and OpenImages.
- 11
- Linear Complexity Self-Attention with $3^{\text{rd}}$ Order Polynomials11
Two modules are proposed that can be used as ”drop-in” replacements for more-complex non-local and self-attention layers in state-of-the-art CNNs and ViT architectures and achieve comparable, if not better, performance across a wide range of computer vision tasks while keeping a complexity equivalent to a standard linear layer.
- Low rank driven robust facial landmark regression11
Extensive results on several wild benchmarks such as COFW, LFPW and Helen demonstrate that the proposed method is robust to facial occlusions, pose variations and exaggerated facial expressions.
- Monocular Identity-Conditioned Facial Reflectance Reconstruction10
This paper learns the reflectance prior in image space rather than UV space rather than UV space and presents a framework named ID2Reflectance, which can directly estimate the reflectance maps of a single image while using limited reflectance data for training.
- 10
- 10
- Unified Physical-Digital Attack Detection Challenge9
This paper comprehensively reviews the challenge, detailing the dataset introduction, protocol definition, evaluation criteria, and a summary of published results, and offers potential directions for unified physical-digital attack detection inspired by this competition.
- RealSyn : An Effective and Scalable Multimodal Interleaved Document Transformation Paradigm8
RealSyn is constructed, a dataset combining realistic and synthetic texts, available in three scales: 15M, 30M, and 100M, which significantly enhances contrastive vision-language representation learning and demonstrates robust scalability.
- Interact2Ar: Full-Body Human-Human Interaction Generation via Autoregressive Diffusion Models8
The first end-to-end text-conditioned autoregressive diffusion model for generating full-body, human-human interactions, Interact2Ar is introduced, the first end-to-end text-conditioned autoregressive diffusion model for generating full-body, human-human interactions.
- 8
- 8
- OVGaussian: Generalizable 3D Gaussian Segmentation with Open Vocabularies8
Experimental results demonstrate that OVGaussian significantly outperforms baseline methods, exhibiting robust cross-scene, cross-domain, and novel-view generalization capabilities.
- Video Face Editing Using Temporal-Spatial-Smooth Warping8
A temporal-spatial-smooth warping method that is robust to the subtly temporal incoherence of the facial feature point localizations and is effective to preserve the temporal coherence and spatial smoothness of the control lattices for editing faces in videos is proposed.
- 8
- 7
- LMAD: Integrated End-to-End Vision-Language Model for Explainable Autonomous Driving7
This work proposes a novel vision-language framework tailored for autonomous driving, called LMAD, which emulates modern end-to-end driving paradigms by incorporating comprehensive scene understanding and a task-specialized structure with VLMs.
- 7
- Informed Non-Convex Robust Principal Component Analysis With Features7
A novel, elegant, non-convex optimization approach is proposed to decompose a given observation matrix into a low-rank core and the corresponding sparse residual and it is demonstrated that this method is the first to wholly harness the power of non- Convexity over convexity in terms of both recoverability and speed.
- 6
- 6
- Egocentric Co-Pilot: Web-Native Smart-Glasses Agents for Assistive Egocentric AI5
Egocentric Co-Pilot is presented, a web-native neuro-symbolic framework that runs on smart glasses and uses a Large Language Model (LLM) to orchestrate a toolbox of perception, reasoning, and web tools that support education, accessibility, and social inclusion for people who may benefit most from contextual, egocentric AI.
- 5
- Hierarchical Convolutional Neural Network for Face Detection5
The proposed approach of hierarchical convolutional neural network for face detection reaches the state-of-the-art performance and is evaluated on the AFW dataset, FDDB dataset and Pascal Faces, and it has effective generalization.
- UniForce: A Unified Latent Force Model for Robot Manipulation with Diverse Tactile Sensors4
UniForce is proposed, a novel unified tactile representation learning framework that learns a shared latent force space across diverse tactile sensors, and can be plugged into downstream force-aware robot manipulation tasks with zero-shot transfer, without retraining or finetuning.
- ImHead: A Large-Scale Implicit Morphable Model for Localized Head Modeling4
ImHead is proposed, a novel implicit 3DMM that not only models expressive 3D head avatars but also facilitates localized editing of the facial features, and provides an interpretable solution for 3D face manipulation, allowing the user to make localized edits.
- RetouchLLM: Training-free Code-based Image Retouching with Vision Language Models4
RetouchLLM is proposed, a training-free white-box image retouching system, which requires no training data and performs interpretable, code-based retouching directly on high-resolution images.
- 4
- S3-Face: SSS-Compliant Facial Reflectance Estimation via Diffusion Priors3
This paper introduces the use of Stable Diffusion as a prior for facial reflectance estimation, achieving robust results with minimal captured data for fine-tuning, and presents S3-Face, a comprehensive framework capable of producing SSS-compliant skin reflectance from in-the-wild images.
- 3
- 3
- Rethinking the Domain Gap in Near-infrared Face Recognition3
It is observed that large neural networks, unlike their smaller counterparts, when pretrained on large scale homogeneous VIS data, demonstrate exceptional zero-shot performance in HFR, suggesting that the domain gap might be less pronounced than previously believed.
- High-Fidelity Facial Albedo Estimation via Texture Quantization3
A novel facial albedo reconstruction model, HiFiAlbedo, which recovers the albedo map directly from a single image without the need for captured albedo data and exhibits excellent generalizability and is capable of achieving high-fidelity results for in-the-wild facial albedo recovery.
- 3
- Single-view Image to Novel-view Generation for Hand-Object Interactions2
This work introduces the HO123, a novel method to synthesize novel-view hand-object interaction images from a single image, which significantly outperforms state-of-the-art novel-view synthesis for hand-object interactions.
- Unlocking the Potential of Diffusion Priors in Blind Face Restoration2
A unified network FLIPNET that switches between two modes to resolve specific gaps in face restoration and outperforms previous diffusion prior based BFR methods in terms of authenticity and fidelity and outperforms the naive degradation model in modeling the real-world degradations.
- 2
- 3DGazeNet: Generalizing Gaze Estimation with Weak-Supervision from Synthetic Views2
This work overcome the absence of compatible ground truth by tipping a rigid 3D eyeball template on existing gaze datasets and proposes an automatic pipeline to retrieve robust gaze pseudo-labels from arbitrary face images and design a multi-view supervision framework to balance their effect during training.
- Supervised descent method with low rank and sparsity constraints for robust face alignment2
A sparsity regularized supervised descent model is proposed by enforcing the sparse structure of the descent directions under the l1constraint, which makes the model more effective in computation and robust to partial occlusion.
- Probabilistic modeling of disparity uncertainty for robust and efficient stereo matching1
A new uncertainty-aware stereo matching framework is proposed that systematically analyze data uncertainty based on the probabilistic distribution of disparity and efficiently estimate model uncertainty without repeated model training.
- 1
- 1
- 1
- 1
- Object tracking based on the joint model using L2-norm minimization1
The experimental results finally show that compared with other typical algorithms, the proposed algorithm has stronger robustness in the case of illumination, scale changes, shelter, rotation and so on.
- –
- –
- Color When It Counts: Grayscale-Guided Online Triggering for Always-On Streaming Video Sensing–
The proposed ColorTrigger is an online training-free trigger that selectively activates color capture based on windowed grayscale affinity analysis, designed for real-time edge deployment and enabling practical always-on video sensing on resource-constrained devices.
- –
- De-biasing Facial Albedo Estimation via Visual-Textual Cues–
The relationship between albedo and face attributes is reconsidered and an ID2Albedo++ method to directly estimate albedo without constraining illumination is proposed and outperforms state-of-the-art albedo estimation methods in terms of both accuracy and fidelity.
- Balancing Marker and Markerless Modes in Vision-Based Tactile Sensors with a Translucent Skin–
Results indicate that translucent skin substantially mitigate the traditional trade-off between marker-based and markerless tactile sensing, thereby expanding the applicability of multi-modal VBTS in tactile robotics.
- –
- –
- –
- –
- Dynamic Novel View Synthesis in High Dynamic Range–
HDR-4DGS is introduced, a Gaussian Splatting-based architecture featured with an innovative dynamic tone-mapping module that explicitly connects HDR and LDR domains, maintaining temporal radiance coherence by dynamically adapting tone-mapping functions according to the evolving radiance distributions across the temporal dimension.
- –
- –
- AvatarMMC: 3D Head Avatar Generation and Editing with Multi-Modal Conditioning–
This approach is the first to introduce multi-modal conditioning to 3D avatar generation and editing with multi-modal conditioning based on a 3D Generative Adversarial Network (GAN) and a Latent Diffusion Model (LDM).
- –
- –
- –
- –
- –
- –
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 2026-10-14. 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.