ICPR 2020

695 papers
The 2020 edition of the International Conference on Pattern Recognition (ICPR) included 942 papers, with a strong emphasis on multimodal learning and representation. Research directions involved multimodal feature fusion, fine-grained alignment, and prediction for autonomous agents. Notable attention was given to bridging the domain gap for synthetic training data and mitigating shortcut learning in self-supervised models. Methodological trends included the exploration of neural collapse, geometric feature packing, and permutation invariance in 3D point cloud learning. Technical contributions also addressed neural architecture search for embedded systems and knowledge distillation through automatic student network search. In terms of applications, the proceedings featured work on brain network encoding for disease classification, semi-supervised person re-identification, and image-to-latex conversion using Transformer models. Prolific authors contributing to this year’s research included Rita Cucchiara, Ran He, Simone Calderara, and Edwin R. Hancock. The research collectively focused on improving model robustness against label noise and optimizing cyclic consistency for image-to-image translation.

Top Research Directions

Research directions most represented by ICPR 2020 papers.

Fusing 2D Semantic Features with 3D Geometric Representations 5
Preserving Tiny Features Against Spatial Resolution Degradation 5
Generalizing from Synthetic Data to the Real World 4
Lightweight Deep Learning for Edge Deployment 4
Probabilistic Trajectory Forecasting for Autonomous Systems 3
Composite Loss Functions for Perceptual Image Restoration 3
Cross-Modal Attention Fusion for Heterogeneous Data Integration 3
Dynamic Modality Weighting for Multimodal Fusion 3
Expanding the Receptive Field of Convolutional Networks 3
The Sensitivity of Multi-Objective Loss Balancing 3
Unified Representations via Multimodal Feature Fusion 3
Bypassing the Computational Bottleneck of Adversarial Training 3
Test Time Adaptation for Distribution Shift 3
Mitigating Shortcut Learning in Visual Self Supervision 2
Geometric Saturation in High Cardinality Embedding Spaces 2

Top Authors

AuthorPapers
Alberto Del Bimbo9
Rita Cucchiara9
Ran He8
Hong Liu6
Linlin Shen6
Simone Calderara6
Alessio Del Bue5
Edwin Hancock5
Giovanni Maria Farinella5
Josef Kittler5
Nam Ik Cho5
Stéphane Lathuilière5
Vittorio Murino5
Yongli Hu5
AIHUA ZHENG4
Andreas Dengel4
Antonino Furnari4
Costantino Grana4
Efstratios Gavves4
Elisa Ricci4

Impactful Papers

Papers from ICPR 2020 that shaped research clusters.

Foundational

Influential

Papers

Title
Two-Level Attention-based Fusion Learning for RGB-D Face Recognition
Two-Stage Adaptive Object Scene Flow Using Hybrid CNN-CRF Model
Two-Stream Temporal Convolutional Network for Dynamic Facial Attractiveness Prediction
UCCTGAN: Unsupervised Clothing Color Transformation Generative Adversarial Network
UHRSNet: A Semantic Segmentation Network Specifically for Ultra-High-Resolution Images
Uncertainty-sensitive Activity Recognition: A Reliability Benchmark and the CARING Models
Understanding Integrated Gradients with SmoothTaylor for Deep Neural Network Attribution
Unsupervised Contrastive Photo-to-Caricature Translation based on Auto-distortion
Unsupervised Disentangling of Viewpoint and Residues Variations by Substituting Representations for Robust Face Recognition
Unsupervised Domain Adaptation for Object Detection in Cultural Sites
Unsupervised Domain Adaptation with Multiple Domain Discriminators and Adaptive Self-Training
Unsupervised Face Manipulation via Hallucination
Unsupervised Feature Learning for Event Data: Direct vs Inverse Problem Formulation
Unsupervised Learning of Facial Landmarks based on Inter-Intra Subject Consistencies
Unsupervised Moving Object Detection through Background Models for PTZ Camera
Unveiling Groups of Related Tasks in Multi - Task Learning
User-Independent Gaze Estimation by Extracting Pupil Parameter and Its Mapping to the Gaze Angle
Vacant Parking Space Detection based on Task Consistency and Reinforcement Learning
Variational Capsule Encoder
Variational Deep Embedding Clustering by Augmented Mutual Information Maximization
Variational Inference with Latent Space Quantization for Adversarial Resilience
Verifying the Causes of Adversarial Examples
Vertex Feature Encoding and Hierarchical Temporal Modeling in a Spatio-Temporal Graph Convolutional Network for Action Recognition
Video Analytics Gait Trend Measurement for Fall Prevention and Health Monitoring
Video-based Facial Expression Recognition using Graph Convolutional Networks
Video Lightening with Dedicated CNN Architecture
Video Object Detection Using Object's Motion Context and Spatio-Temporal Feature Aggregation
Video Summarization with a Dual Attention Capsule Network
Visual Localization for Autonomous Driving: Mapping the Accurate Location in the City Maze
Visual Style Extraction from Chart Images for Chart Restyling
VITON-GT: An Image-based Virtual Try-On Model with Geometric Transformations
VOWEL: A Local Online Learning Rule for Recurrent Networks of Probabilistic Spiking Winner- Take-All Circuits
VR Sickness Assessment with Perception Prior and Hybrid Temporal Features
VSB2-Net: Visual-Semantic Bi-Branch Network for Zero-Shot Hashing
VSR++: Improving Visual Semantic Reasoning for Fine-Grained Image-Text Matching
VTT: Long-term Visual Tracking with Transformers
Watch Your Strokes: Improving Handwritten Text Recognition with Deformable Convolutions
Weakly Supervised Attention Rectification for Scene Text Recognition
Weakly Supervised Body Part Segmentation with Pose based Part Priors
Weakly Supervised Geodesic Segmentation of Egyptian Mummy CT Scans
Webly Supervised Image-Text Embedding with Noisy Tag Refinement
What and How? Jointly Forecasting Human Action and Pose
What nodes vote to? Graph classification without readout phase
Zero-Shot Text Classification with Semantically Extended Graph Convolutional Network
Zoom-CAM: Generating Fine-grained Pixel Annotations from Image Labels

« prev Page 14/14