2D-3D Attention and Entropy for Pose Robust 2D Facial Recognition
Not on the map. Why?
Abstract (arXiv)
Despite recent advances in facial recognition, there remains a fundamental issue concerning degradations in performance due to substantial perspective (pose) differences between enrollment and query (probe) imagery. Therefore, we propose a novel domain adaptive framework to facilitate improved performances across large discrepancies in pose by enabling image-based (2D) representations to infer properties of inherently pose invariant point cloud (3D) representations. Specifically, our proposed framework achieves better pose invariance by using (1) a shared (joint) attention mapping to emphasize common patterns that are most correlated between 2D facial images and 3D facial data and (2) a joint entropy regularizing loss to promote better consistencyunicodex2014enhancing correlations among the intersecting 2D and 3D representationsunicodex2014by leveraging both attention maps. This framework is evaluated on FaceScape and ARL-VTF datasets, where it outperforms competitive methods by achieving profile (90unicodex00b0unicodex002b) TAR @ 1unicodex0025 FAR improvements of at least 7.1unicodex0025 and 1.57unicodex0025, respectively.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 57 of its 82 references that are in the map. First posted on arXiv 14 May 2025.
Builds on
57 of 82 references.
References · 57
- 2015Deep Residual Learning for Image Recognitioncited by 52,096
- 2020Denoising Diffusion Probabilistic Modelscited by 21,076
- 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,364
- 2015Distilling the Knowledge in a Neural Networkcited by 10,324
- 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,401
- 2015FaceNet: A Unified Embedding for Face Recognition and Clusteringcited by 2,849
- 2015Learning Transferable Features with Deep Adaptation Networkscited by 1,329
- 2017Adversarial Discriminative Domain Adaptationcited by 1,311
- 2014Depth Map Prediction from a Single Image using a Multi-Scale Deep Networkcited by 1,274
- 2016Deep CORAL: Correlation Alignment for Deep Domain Adaptationcited by 1,200
- 2018CosFace: Large Margin Cosine Loss for Deep Face Recognitioncited by 971
- 2016Unsupervised Monocular Depth Estimation with Left-Right Consistencycited by 820
38 more
- 2018PointCNN: Convolution On X-Transformed Pointscited by 764
- 2017SphereFace: Deep Hypersphere Embedding for Face Recognitioncited by 737
- 2018Taskonomy: Disentangling Task Transfer Learningcited by 634
- 2017Generative Adversarial Networks: An Overviewcited by 590
- 2022AdaFace: Quality Adaptive Margin for Face Recognitioncited by 391
- 2022CrossPoint: Self-Supervised Cross-Modal Contrastive Learning for 3D Point Cloud Understandingcited by 198
- 2020FaceScape: a Large-scale High Quality 3D Face Dataset and Detailed Riggable 3D Face Predictioncited by 196
- 2019Forward and Backward Information Retention for Accurate Binary Neural Networkscited by 169
- 20222DPASS: 2D Priors Assisted Semantic Segmentation on LiDAR Point Cloudscited by 158
- 2019xMUDA: Cross-Modal Unsupervised Domain Adaptation for 3D Semantic Segmentationcited by 129
- 20183DMV: Joint 3D-Multi-View Prediction for 3D Semantic Scene Segmentationcited by 118
- 2016Perspective Transformer Nets: Learning Single-View 3D Object Reconstruction without 3D Supervisioncited by 106
- 2019DADA: Depth-aware Domain Adaptation in Semantic Segmentationcited by 97
- 2021Bidirectional Projection Network for Cross Dimension Scene Understandingcited by 95
- 2023IDiff-Face: Synthetic-based Face Recognition through Fizzy Identity-Conditioned Diffusion Modelscited by 72
- 2018Learning Shape Priors for Single-View 3D Completion and Reconstructioncited by 68
- 2021Pri3D: Can 3D Priors Help 2D Representation Learning?cited by 61
- 2021Learning 3D Semantic Segmentation with only 2D Image Supervisioncited by 60
- 2021SimIPU: Simple 2D Image and 3D Point Cloud Unsupervised Pre-Training for Spatial-Aware Visual Representationscited by 55
- 20213D-to-2D Distillation for Indoor Scene Parsingcited by 48
- 20173D Face Morphable Models "In-the-Wild"cited by 44
- 2021Image2Point: 3D Point-Cloud Understanding with 2D Image Pretrained Modelscited by 41
- 2014Effective Face Frontalization in Unconstrained Imagescited by 34
- 2022AT-DDPM: Restoring Faces degraded by Atmospheric Turbulence using Denoising Diffusion Probabilistic Modelscited by 34
- 2018Pose-Robust Face Recognition via Deep Residual Equivariant Mappingcited by 28
- 2017Multi-Task Convolutional Neural Network for Pose-Invariant Face Recognitioncited by 24
- 2022Let Images Give You More:Point Cloud Cross-Modal Training for Shape Analysiscited by 24
- 2022Learning 3D Scene Priors with 2D Supervisioncited by 24
- 2021A Large-Scale, Time-Synchronized Visible and Thermal Face Datasetcited by 23
- 2023Controllable Inversion of Black-Box Face Recognition Models via Diffusioncited by 22
- 2020Self-supervised Feature Learning by Cross-modality and Cross-view Correspondencescited by 19
- 2016Do We Really Need to Collect Millions of Faces for Effective Face Recognition?cited by 17
- 2023Mask3D: Pre-training 2D Vision Transformers by Learning Masked 3D Priorscited by 15
- 2020Cross-Domain Identification for Thermal-to-Visible Face Recognitioncited by 14
- 2023PointMBF: A Multi-scale Bidirectional Fusion Network for Unsupervised RGB-D Point Cloud Registrationcited by 14
- 2017Generative Adversarial Network-based Synthesis of Visible Faces from Polarimetric Thermal Facescited by 11
- 2017Deep Convolutional Neural Network for Age Estimation based on VGG-Face Modelcited by 11
- 2020Multi-Scale Thermal to Visible Face Verification via Attribute Guided Synthesiscited by 11
50 most cited of 57 All 57 (JSON)
Built on by
Citing papers per year: 2026: 1.
1 paper.