Computer Vision - ECCV 2022 Workshops

Name: Computer Vision - ECCV 2022 Workshops | Tel Aviv, Israel, October 23-27, 2022, Proceedings, Part VI
Brand: Springer
Price: 106.99 EUR
Availability: OnlineOnly

Tel Aviv, Israel, October 23-27, 2022, Proceedings, Part VI

Leonid Karlinsky Tomer Michaeli Ko Nishino(Herausgeber*in)

Springer (Verlag)

Erschienen am 18. Februar 2023

XXVI, 783 Seiten

E-Book

PDF mit Wasserzeichen-DRM

Systemvoraussetzungen

978-3-031-25075-0 (ISBN)

106,99 €inkl. 7% MwSt.

Systemvoraussetzungen

für PDF mit Wasserzeichen-DRM

E-Book Einzellizenz

Als Download verfügbar

Beschreibung

The 8-volume set, comprising the LNCS books 13801 until 13809, constitutes the refereed proceedings of 38 out of the 60 workshops held at the 17th European Conference on Computer Vision, ECCV 2022. The conference took place in Tel Aviv, Israel, during October 23-27, 2022; the workshops were held hybrid or online.

The 367 full papers included in this volume set were carefully reviewed and selected for inclusion in the ECCV 2022 workshop proceedings. They were organized in individual parts as follows:

Part I: W01 - AI for Space; W02 - Vision for Art; W03 - Adversarial Robustness in the Real World; W04 - Autonomous Vehicle Vision

Part II: W05 - Learning With Limited and Imperfect Data; W06 - Advances in Image Manipulation;

Part III: W07 - Medical Computer Vision; W08 - Computer Vision for Metaverse; W09 - Self-Supervised Learning: What Is Next?;

Part IV: W10 - Self-Supervised Learning for Next-Generation Industry-LevelAutonomous Driving; W11 - ISIC Skin Image Analysis; W12 - Cross-Modal Human-Robot Interaction; W13 - Text in Everything; W14 - BioImage Computing; W15 - Visual Object-Oriented Learning Meets Interaction: Discovery, Representations, and Applications; W16 - AI for Creative Video Editing and Understanding; W17 - Visual Inductive Priors for Data-Efficient Deep Learning; W18 - Mobile Intelligent Photography and Imaging;

Part V: W19 - People Analysis: From Face, Body and Fashion to 3D Virtual Avatars; W20 - Safe Artificial Intelligence for Automated Driving; W21 - Real-World Surveillance: Applications and Challenges; W22 - Affective Behavior Analysis In-the-Wild;

Part VI : W23 - Visual Perception for Navigation in Human Environments: The JackRabbot Human Body Pose Dataset and Benchmark; W24 - Distributed Smart Cameras; W25 - Causality in Vision; W26 - In-Vehicle Sensing and Monitorization; W27 - Assistive Computer Vision and Robotics; W28 - Computational Aspectsof Deep Learning;

Part VII: W29 - Computer Vision for Civil and Infrastructure Engineering; W30 - AI-Enabled Medical Image Analysis: Digital Pathology and Radiology/COVID19; W31 - Compositional and Multimodal Perception;

Part VIII: W32 - Uncertainty Quantification for Computer Vision; W33 - Recovering 6D Object Pose; W34 - Drawings and Abstract Imagery: Representation and Analysis; W35 - Sign Language Understanding; W36 - A Challenge for Out-of-Distribution Generalization in Computer Vision; W37 - Vision With Biased or Scarce Data; W38 - Visual Object Tracking Challenge.

Weitere Details

Weitere Ausgaben

Inhalt

Intro
Foreword
Preface
Organization
Contents - Part VI
W22 - Competition on Affective Behavior Analysis In-the-Wild
W22 - Competition on Affective Behavior Analysis In-the-Wild
Geometric Pose Affordance: Monocular 3D Human Pose Estimation with Scene Constraints
1 Introduction
2 Related Work
3 Geometric Pose Affordance Dataset (GPA)
4 Geometry-Aware Pose Estimation
4.1 Pose Estimation Baseline Model
4.2 Geometric Consistency Loss and Encoding
4.3 Overall Training
5 Experiments
5.1 Baselines
5.2 Effectiveness of Geometric Affordance
6 Discussion and Conclusion
References
Affective Behaviour Analysis Using Pretrained Model with Facial Prior
1 Introduction
2 Related Work
3 Proposed Method
3.1 Formulation
3.2 EMMA
3.3 Masked CoTEX
4 Experiment
4.1 Benchmarks and Evaluation Metrics
4.2 Training Details
4.3 Recognition Results
4.4 Ablation Study
5 Conclusions
References
Facial Affect Recognition Using Semi-supervised Learning with Adaptive Threshold
1 Introduction
2 Method
2.1 MFAR: Multi-task Facial Affect Recognition
2.2 SS-MFAR: Semi-supervised Multi-task Facial Affect Recognition
2.3 Problem Formulation
2.4 Adaptive Threshold
2.5 Supervision Loss
2.6 Unsupervised and Consistency Loss
2.7 Overall Loss
3 Dataset and Implementation Details
3.1 Dataset
3.2 Implementation Details
3.3 Evaluation Metrics
4 Results and Discussion
4.1 Performance on Validation Set
4.2 Ablation Studies
4.3 Performance on Test Set
5 Conclusions
References
MT-EmotiEffNet for Multi-task Human Affective Behavior Analysis and Learning from Synthetic Data
1 Introduction
2 Proposed Approach
2.1 Multi-task Learning Challenge
2.2 Learning from Synthetic Data Challenge
3 Experimental Study
3.1 FER for Static Images
3.2 Multi-task-Learning Challenge
3.3 Learning from Synthetic Data Challenge
4 Conclusions
References
Ensemble of Multi-task Learning Networks for Facial Expression Recognition In-the-Wild with Learning from Synthetic Data
1 Introduction
2 Method
2.1 Data Pre-processing
2.2 Model Architecture
3 Experiments and Results
3.1 Training Setup
3.2 Performance Evaluation
4 Conclusion
References
PERI: Part Aware Emotion Recognition in the Wild
1 Introduction
2 Related Work
3 Method
3.1 MediaPipe Holistic Model
3.2 The Emotic Model
3.3 Part Aware Spatial Image
3.4 Context Infusion Blocks
4 Experiments
4.1 Experiment Setup
4.2 Quantitative Results
4.3 Qualitative Results
4.4 Ablation Study
5 Conclusion
References
Facial Expression Recognition with Mid-level Representation Enhancement and Graph Embedded Uncertainty Suppressing
1 Introduction
2 Related Work
3 Method
3.1 Overview
3.2 Mid-level Representation Enhancement
3.3 Graph Embedded Uncertainty Suppressing
3.4 Latent Distribution Mining and Pairwise Uncertainty Estimation
3.5 Mixed Loss Functions
3.6 Prediction Merging and Correcting
4 Experiments
4.1 Dataset
4.2 Settings
4.3 Results
5 Conclusion
References
Deep Semantic Manipulation of Facial Videos
1 Introduction
2 Related Work
3 Proposed Method
3.1 AI Tool
3.2 Person-Specific Dataset of Facial Expressions
3.3 Expression Decoder Network
3.4 Synthesis of Photorealistic Manipulated videos
4 Experimental Results
4.1 Qualitative Evaluation
4.2 Quantitative Evaluation
4.3 User Studies
5 Conclusion and Future Work
References
BYEL: Bootstrap Your Emotion Latent
1 Introduction
2 Related Works
2.1 Self-supervised Representation Learning
2.2 Human Emotion Analysis
3 Problem Description
4 Method
4.1 Bootstrap Your Emotion Latent
4.2 Transfer-Learning
4.3 Loss
5 Experiments
5.1 Dataset
5.2 Settings
5.3 Metric
5.4 Results
6 Conclusions
References
Affective Behavior Analysis Using Action Unit Relation Graph and Multi-task Cross Attention
1 Introduction
2 Related Work
3 Proposed Method
3.1 AU Relation Graph
3.2 FER and VA Estimation Heads
3.3 Loss Function
4 Experiment
4.1 Dataset
4.2 Evaluation Metrics
4.3 Experiment Setting
4.4 Result
5 Conclusion
References
Multi-Task Learning Framework for Emotion Recognition In-the-Wild
1 Introduction
2 Related Works
3 Method
3.1 Features
3.2 Temporal Encoder
3.3 Single Task Loss Function
3.4 Multi-Task Learning Framework
4 Experiments
4.1 Dataset
4.2 Experiment Setup
4.3 Overall Results on the Validation Set
4.4 Model Ensemble
4.5 Results on the Test Set
5 Conclusion
References
ABAW: Learning from Synthetic Data & Multi-task Learning Challenges
1 Introduction
2 Competition Corpora
2.1 Multi-task Learning Challenge
2.2 Learning from Synthetic Data Challenge
3 Evaluation Metrics for Each Challenge
3.1 Multi-task Learning Challenge
3.2 Learning from Synthetic Data Challenge
4 Baseline Networks and Performance
4.1 Multi-task Learning Challenge
4.2 Learning from Synthetic Data Challenge
5 Conclusion
References
Two-Aspect Information Interaction Model for ABAW4 Multi-task Challenge
1 Introduction
2 Methodology
2.1 ROI Feature Extraction
2.2 Interaction Module
2.3 Temporal Smoothing and Classifier Module
2.4 Losses
3 Experiments
3.1 Datasets
3.2 Training Details
3.3 Evaluation Metric
3.4 Baseline Model
4 Results
4.1 Comparison with Baseline Models
4.2 Different Temporal Smoothing Methods
4.3 Different Backbones
4.4 Different Semantic Information
4.5 Detail Results
5 Conclusions
References
Facial Expression Recognition In-the-Wild with Deep Pre-trained Models
1 Introduction
2 Related Work
3 Method
4 Experimental Results
5 Conclusion
References
W23 - Visual Perception for Navigation in Human Environments: The JackRabbot Human Body Pose Dataset and Benchmark
W23 - Visual Perception for Navigation in Human Environments: The JackRabbot Human Body Pose Dataset and Benchmark
Robustness of Embodied Point Navigation Agents
1 Introduction
2 Related Work
3 Method
3.1 Habitat Environment
3.2 PointNav Task
3.3 Metrics
3.4 Visual Corruptions
3.5 Dynamics Corruptions
4 Experimental Setup
4.1 Agents
4.2 Dataset
4.3 Configurations
5 Results
6 Conclusion and Limitations
References
W24 - Distributed Smart Cameras
W24 - Distributed Smart Cameras
CounTr: An End-to-End Transformer Approach for Crowd Counting and Density Estimation
1 Introduction
2 Related Works
2.1 Crowd Counting and Density Estimation
2.2 Transformers in Vision
3 Methodology
3.1 Preliminaries on Transformers for Crowd Counting
3.2 The CounTr Model
4 Experiments
4.1 Implementation Details and Datasets
4.2 Results and Discussion
4.3 Ablation Study
4.4 Visualization on Density Maps
5 Conclusion
References
On the Design of Privacy-Aware Cameras: A Study on Deep Neural Networks
1 Introduction
2 Related Work
3 Experimental Study
3.1 Experimental Setup
3.2 Semantic Segmentation Sensitivity to Color and Defocus Blur
3.3 LPDR Sensitivity to Defocus Blur
4 Conclusions
References
RelMobNet: End-to-End Relative Camera Pose Estimation Using a Robust Two-Stage Training
1 Introduction
2 Related Work
2.1 Feature Correspondence Methods
2.2 End-to-End Methods
3 Methodology
3.1 Dataset Image Pairs
3.2 Architecture Details
4 Experiments and Analysis
4.1 Baseline Comparison
4.2 Two-Stage vs One-Stage Approach
4.3 Secondary Data Collection and Evaluation
5 Conclusion
References
Cross-Camera View-Overlap Recognition
1 Introduction
2 Related Work
3 Decentralising the Recognition of View Overlaps
3.1 Camera Messaging
3.2 Sharing Query Views
3.3 View-Overlap Recognition
4 Validation
4.1 Strategies Under Comparison
4.2 Implementation Details
4.3 Dataset and View-Overlap Annotation
4.4 Performance Measures
4.5 Parameter Setting
4.6 Discussion
5 Conclusion
References
Activity Monitoring Made Easier by Smart 360-degree Cameras
1 Introduction
2 Related Work
3 Equirectangular Images and Their Geometry
3.1 From Equirectangular Image to Ground-Plane Coordinates
4 Activity Monitoring in 360-degree Camera Network
4.1 Correspondences Between the Camera Reference Systems
4.2 Detecting Activities Within Areas of Interests
4.3 Detecting Areas of Interest in the Scene
5 Implementation Details
6 Experiments
6.1 Finding Correspondences Between Camera Reference Systems
6.2 Activity Detection Within RoI
6.3 Detecting Areas of Interest
7 Conclusions and Future Work
References
Seeing Objects in Dark with Continual Contrastive Learning
1 Introduction
2 Proposed Method
3 Related Work and Experiments
3.1 Ablation Studies, and Detailed Hyperparameter Sensitivity Analysis of Our Method
3.2 Comparison with State-of-the-Art
4 Conclusion
References
Towards Energy-Efficient Hyperspectral Image Processing Inside Camera Pixels
1 Introduction
2 Related Work
2.1 Energy-Efficient On-Device Vision
2.2 HSI Camera Operations
2.3 CNN Models for HSI
3 Algorithm-Hardware Co-design for HSI
3.1 Preliminaries
3.2 Scheduling PIP Computation for HSI
3.3 PIP-Optimized HSI Models
3.4 PIP-Enabled Compression
4 Experimental Results
4.1 Experimental Setup
4.2 Quantification of Compression
4.3 Classification Accuracy
4.4 Analysis of Energy-Efficiency
5 Summary and Conclusions
References
Identifying Auxiliary or Adversarial Tasks Using Necessary Condition Analysis for Adversarial Multi-task Video Understanding*-4pt
1 Introduction
2 Related Work
3 Dataset and Necessary Condition Analysis
3.1 HVU Dataset
3.2 Necessary Condition Analysis (NCA)
3.3 Human Study/Necessity Annotation
3.4 Scene-Invariant Splits of the HVU Dataset
3.5 Evaluation Metric for Scene Bias
4 Method
5 Experiments
5.1 Implementation Details
5.2 Synthetic Data Experiments
5.3 HVU Dataset Experiments
6 Conclusion
References
Deep Multi-modal Representation Schemes for Federated 3D Human Action Recognition
1 Introduction
2 Related Work
3 3D Action Recognition Dataset
4 Single-Modality Analysis
5 Multi-modal Analysis
6 Experimental Results
7 Conclusions
References
Spatio-Temporal Attention for Cloth-Changing ReID in Videos
1 Introduction
2 Related Work
2.1 Short-Term Cloth-Consistent ReID
2.2 Video-Based ReID
2.3 Long-Term Cloth-Changing ReID
3 Methodology
3.1 Vision Transformer(ViT)-Based Baseline
3.2 Body-Shape-Motion Features Using VIBE
3.3 Our Proposed Framework: DST-Base with VIBE
4 Experiments
4.1 Datasets
4.2 Comparison with State-of-the-art Methods
4.3 Ablation Study on Our Baseline-VIBE Hybrid Model
4.4 Ablation Study with Different Tracklet Length
5 Conclusions
References
W25 - Causality in Vision
W25 - Causality in Vision
Towards Interpreting Computer Vision Based on Transformation Invariant Optimization
1 Introduction
2 Method
2.1 Iterative Rotation Algorithm
2.2 Gray Scale Initialization
3 Experiments
3.1 Visualization of Different Classes
3.2 Visualization of Different Classes
3.3 Debugging of Models According Visualization
4 Conclusions
References
Investigating Neural Network Training on a Feature Level Using Conditional Independence
1 Introduction
2 Related Work
3 Methodology
3.1 Conditional Independence Tests
4 Selected Classification Tasks and Feature Sets
4.1 Toy Example
4.2 MNIST
4.3 ISIC-archive
5 Experiments
5.1 MLP Learning a Toy Example
5.2 CNN Learning MNIST Three Versus Eight
5.3 Modern Architecture Learning Melanoma Classification
6 Conclusions
References
Deep Structural Causal Shape Models
1 Introduction
2 Related Work
3 Deep Structural Causal Shape Models
3.1 Counterfactual Inference of High-Dimensional Variables
3.2 Implementing Causal Shape Models
3.3 Convolutional Mesh Conditional Variational Autoencoder
4 Experiments
4.1 Setup
4.2 Associational Level: Shape Reconstruction
4.3 Interventional Level: Population-Level Shape Generation
4.4 Counterfactual Level: Subject-Specific Shape Generation
5 Conclusion
Appendix A Additional Background Material
A.1 DSCM Mechanisms
A.2 Pearl's Causal Hierarchy
A.3 Counterfactual Inference using DSCM Framework
Appendix B Implementation of Causal Shape Models
B.1 Low-Dimensional Covariate Mechanisms
B.2 ELBO full proof
Appendix C Training Causal Shape Models
Appendix D Further Experiments with Brain Stem Meshes
D.1 Association
D.2 Population-Level
D.3 Subject-Specific
Appendix E Preliminary Results on Hippocampus Meshes
References
NICO Challenge: Out-of-Distribution Generalization for Image Recognition Challenges
1 Introduction
2 Related Work
3 NICO++ Dataset
4 Challenge Tracks
4.1 Track 1: Common Context Generalization
4.2 Track 2: Hybrid Context Generalization
5 Pretraining and Model Initialization
6 Conclusion
References
Decoupled Mixup for Out-of-Distribution Visual Recognition
1 Introduction
2 Related Works
3 Method
3.1 Style-Based Decoupled-Mixup
3.2 Context-Aware Decoupled-Mixup
3.3 Frequency-Aware Decoupled-Mixup
3.4 Decoupled-Mixup for OOD Visual Recognition
4 Experiment
4.1 Experiment Settings
4.2 Classification Results
5 Conclusion
References
Bag of Tricks for Out-of-Distribution Generalization
1 Introduction
2 Related Work
2.1 Domain Generalization
2.2 Fine-grained Classification
2.3 Generalization Ability
3 Challenge Description
3.1 Dataset
3.2 Task
4 Method
4.1 Multi-objective Framework Design
4.2 Data Augmentation
4.3 Training Strategy
4.4 Inference Strategy
5 Experiments
5.1 Implementation Details
5.2 Results
5.3 Ablation Studies
6 Conclusions
References
SimpleDG: Simple Domain Generalization Baseline Without Bells and Whistles
1 Introduction
1.1 Domain Generalization
1.2 DomainBed
1.3 NICO++
2 SimpleDG
2.1 Evaluation Metric
2.2 Key Designs of SimpleDG
2.3 More Implementation Detail
2.4 Public Results
2.5 Private Results
3 Conclusion
References
A Three-Stage Model Fusion Method for Out-of-Distribution Generalization
1 Introduction
2 Related Work
3 Method
3.1 Model Architecture
3.2 Loss Functions
3.3 Data Augmentation
4 Experiment
4.1 Dataset
4.2 Data Preprocessing
4.3 Training Details
4.4 Ablation Experiments
4.5 Model Fusion
5 Discussion
6 Conclusion
References
Bootstrap Generalization Ability from Loss Landscape Perspective
1 Introduction
2 Related Work
3 Method
3.1 Overview of Our Training Paradigm
3.2 Adaptive Learning Rate Scheduler
3.3 Supervised Self-distillation
3.4 Architecture
3.5 Regularization
4 Experiments
4.1 Datasets
4.2 Settings
4.3 Domain Generalization Results on NICO++ Dataset
4.4 Leave-One-Domain-Out Results on PACS Dataset
4.5 Leave-One-Domain-Out Results on VLCS Dataset
4.6 Visualizations
5 Conclusion
References
Domain Generalization with Global Sample Mixup
1 Introduction
2 Related Works
2.1 Image Classification
2.2 Domain Adaptation and Generalization
3 Methodology
3.1 Problem Settings
3.2 Global Sample Mixup
3.3 Data Augmentation
3.4 Training Strategy
3.5 Model Ensemble
4 Experiments
4.1 Implementation Details
4.2 Single Model with Different Backbones
4.3 Resizing and Cropping Strategies
4.4 Ablation Study
5 Conclusions
6 Future Work
References
Meta-Causal Feature Learning for Out-of-Distribution Generalization
1 Introduction
2 Related Works
2.1 OOD Generalization
2.2 Causal Inference
2.3 Meta Learning
3 Relation Between BMCL, CBAM and CaaM
3.1 From Attentive Feature Learning to Causal Learning
3.2 The Enhanced Causal Learning by BMCL
4 Our Approach
4.1 Overview
4.2 Balanced Task Generation (BTG) Module
4.3 Meta-Causal Feature Learning (MCFL) Module
4.4 Training Strategy
5 Experiments
5.1 Datasets
5.2 Evaluation Protocol
5.3 Implementation Details
5.4 Performance Comparison
5.5 Ablation Study
5.6 In Depth Analysis of Balancing Strategy and Meta Learner
5.7 Case Study
6 Conclusions
References
W26 - In-vehicle Sensing and Monitorization
W26 - In-vehicle Sensing and Monitorization
Detecting Driver Drowsiness as an Anomaly Using LSTM Autoencoders
1 Introduction
2 Related Work
3 The Dataset
4 Methodology
5 Experiment Details
6 Results
7 Conclusion
References
Semi-automatic Pipeline for Large-Scale Dataset Annotation Task: A DMD Application
1 Introduction
2 DMD: Driver Monitoring Dataset
3 Related Work
4 Methodology
4.1 Semi-Automatic Annotation Implementation (SAA)
4.2 Dummy Model
4.3 Automatic Label Assessment (ALA) Strategy
5 Results
6 Discussion
7 Conclusion
8 Future Work
References
Personalization of AI Models Based on Federated Learning for Driver Stress Monitoring
1 Introduction
2 Proposed Approach: Driver-Personalized Federated Learning
3 Experiments
3.1 Dataset and Annotations
3.2 Methodology and Implementation Details
3.3 Evaluation
4 Results and Discussion
5 Conclusion
References
W27 - Assistive Computer Vision and Robotics
W27 - Assistive Computer Vision and Robotics
Multi-scale Motion-Aware Module for Video Action Recognition
1 Introduction
2 Related Work
2.1 Optical Flow Driven Methods
2.2 Exploiting the Relation Between Pixels
2.3 Multi-scale Solutions
3 Methodology
3.1 Notations
3.2 Correlation Operation and Motion-Aware Module
3.3 Multi-Scale Motion-Aware Module
3.4 Installing MSMA on TSM ResNet-50
4 Experiments
4.1 Setup
4.2 Main Results
4.3 Inference Time Comparison with Correlation Methods
4.4 Ablation Study
5 Conclusions
References
Detect and Approach: Close-Range Navigation Support for People with Blindness and Low Vision
1 Introduction
2 Related Works
3 Methods
3.1 Trajectory Estimation
3.2 Object Localization
4 Experiments
4.1 Experiment with the KITTI Dataset
4.2 Simulated Navigation Experiments
4.3 Run Time Analysis
5 Conclusions
References
Multi-modal Depression Estimation Based on Sub-attentional Fusion*-2pt
1 Introduction
2 Related Work
3 Model
3.1 Sub-attentional ConvBiLSTM
3.2 Attentional Fusion Layer
3.3 MUSDL
3.4 SAM
4 Experiments
4.1 Dataset
4.2 Effectiveness of Different Fusion Methods
4.3 Ablation Studies
4.4 Automatic Depression Estimation
5 Conclusion
References
Interactive Multimodal Robot Dialog Using Pointing Gesture Recognition
1 Introduction
2 Related Work
2.1 Natural Language Understanding
2.2 Pointing Recognition
2.3 Combination of NLU and Pointing Recognition
3 Methodology
3.1 Pointing Recognition
3.2 Multimodal Natural Language Understanding
3.3 Interactive Pointing Dialog
4 Dataset
4.1 Human Pointing Dataset
4.2 Natural Language Understanding Dataset
5 Evaluation
5.1 Pointing Recognition
5.2 Multimodal Natural Language Understanding
5.3 End-to-End Results
6 Conclusion and Discussion
References
Cross-Domain Representation Learning for Clothes Unfolding in Robot-Assisted Dressing
1 Introduction
2 Related Work
2.1 Robotic Clothes Unfolding
2.2 Robot-Assisted Dressing
3 Cross-Domain Representation Learning
4 Data Acquisition
5 Experiments and Results
5.1 Experimental Setup
5.2 Approach Evaluation
5.3 Robotic Clothes Unfolding and Assistive Dressing
6 Conclusion
References
Depth-Based In-Bed Human Pose Estimation with Synthetic Dataset Generation and Deep Keypoint Estimation
1 Introduction
2 Related Work
2.1 RGB Image-Based Human Pose Estimation
2.2 Depth-Based Human Pose Estimation
2.3 Sensor-Based Pose Estimation
3 Synthetic Dataset Generation
3.1 Outline of Synthetic Dataset Generation
3.2 Cloth Simulation
3.3 Dataset Details
4 Experiments with Synthetic Dataset
4.1 Training
4.2 Evaluation Metrics
4.3 Experimental Results
5 Experiments Wit Real Scene Dataset
5.1 Aquisition of Real Scene Dataset
5.2 Experimental Results
6 Conclusions and Discussion
References
Matching Multiple Perspectives for Efficient Representation Learning
1 Introduction
2 Related Work
2.1 Self-supervised Learning
2.2 Positive Mining
3 Multi-perspective Views in Self-supervised Learning
3.1 Simple Contrastive Learning Framework and Variants
3.2 Polygon Matching
4 Use-Case: Robotic Vacuum Agent
5 Experiments
5.1 Implementations Details
5.2 Results
6 Conclusion
References
LocaliseBot: Multi-view 3D Object Localisation with Differentiable Rendering for Robot Grasping
1 Introduction
1.1 3D Model Fitting
2 Related Work
2.1 3D Pose Estimation
2.2 Differentiable Rendering
3 Method
3.1 Refinement
4 Experiments
4.1 Evaluation Metrics
4.2 Dataset
4.3 Setup
4.4 Results
4.5 Object Pose Estimation for Grasping
5 Conclusion
References
Fused Multilayer Layer-CAM Fine-Grained Spatial Feature Supervision for Surgical Phase Classification Using CNNs
1 Introduction
2 Related Work
3 Proposed Method
3.1 Training Methodology
3.2 Spatial Context Aware Combined Loss Function
3.3 Inference Methodology
4 Dataset
5 Experimental Results and Analysis
6 Conclusion
References
Representation Learning for Point Clouds with Variational Autoencoders
1 Introduction
2 Learning Point Cloud Generation
2.1 Related Work
2.2 -VAE with Hyperparameters
3 Proposed Method
3.1 Network Architecture and Training
3.2 Hyperparameter Tuning
3.3 Robustness Tests
4 Conclusions
References
Tele-EvalNet: A Low-Cost, Teleconsultation System for Home Based Rehabilitation of Stroke Survivors Using Multiscale CNN-ConvLSTM Architecture
1 Introduction
2 Data Set
3 Tele-EvalNet
3.1 Live Motion Feedback Module
3.2 Deep Learning Based Movement Quality Assessment Model
4 Results and Discussion
4.1 Live Motion Feedback Module
4.2 Setup
4.3 Hyperparameter Optimization
4.4 Comparative Result Analysis
4.5 Ablation Study
5 Conclusion
References
Towards the Computational Assessment of the Conservation Status of a Habitat
1 Introduction
2 Related Work
3 Methods
3.1 Estimation of Green Areas
3.2 Plant Classification
3.3 Detecting a Plant in the Habitat
4 Results
5 Conclusion
References
Augmenting Simulation Data with Sensor Effects for Improved Domain Transfer
1 Introduction
2 Related Works
3 Approach
3.1 Simulator
3.2 Sensor Effects Modeling
3.3 Dataset
3.4 Grounded Language Learning
3.5 User Studies
4 Results
4.1 Model Performance
4.2 User Study
5 Conclusion
6 Future Work
References
Author Index

Systemvoraussetzungen

Als PDF speichern Als Link merken