CVPR2022論文サマリ

ObjectFormer for Image Manipulation Detection and Localization
by: 岡本大和（LINE Computer Vision Lab）
Attetion Manipulation Detection
Proactive Image Manipulation Detection
by: 岡本大和（LINE Computer Vision Lab）
Adversarial examples Attetion Manipulation Detection
Detector-Free Weakly Supervised Group Activity Recognition
by: Chihiro Nakatani（中谷 千洋）
Action recognition Attetion Recognition Video
Self-Supervised Learning of Adversarial Example: Towards Good Generalizations for Deepfake Detection
by: 岡本大和（LINE Computer Vision Lab）
Adversarial examples Fake
How Good Is Aesthetic Ability of a Fashion Model?
by: Anonymous
Dataset
Few-Shot Incremental Learning for Label-to-Image Translation
by: Yoshi Truong
GAN N-shot learning
Active Learning for Open-Set Annotation
by: Anonymous
Active learning
Sparse Instance Activation for Real-Time Instance Segmentation
by: Yoshiki Kubotani
Instance segmentation Segmentation Real-time
Evaluation-Oriented Knowledge Distillation for Deep Face Recognition
by: 鈴木共生
Knowledge distillation Recognition Face
Self-Supervised Video Transformer
by: 志田　遥飛
Video
HyperStyle: StyleGAN Inversion With HyperNetworks for Real Image Editing
by: Shion Hinda
GAN
Optimal Correction Cost for Object Detection Evaluation
by: Masanori YANO
Object detection
NICE-SLAM: Neural Implicit Scalable Encoding for SLAM
by: shoji sonoyama
3D reconstruction Pose estimation SLAM NeRF
Equivariant Point Cloud Analysis via Learning Orientations for Message Passing
by: Naoya Chiba
3D Disentanglement Point cloud
Syntax-Aware Network for Handwritten Mathematical Expression Recognition
by: Atsuki Osanai
Recognition Text Handwritten
Replacing Labeled Real-Image Datasets With Auto-Generated Contours
by: Anonymous
Dataset Representation learning Self supervised learning
DiffusionCLIP: Text-Guided Diffusion Models for Robust Image Manipulation
by: Yoshi Truong
Diffusion
Pushing the Performance Limit of Scene Text Recognizer Without Human Annotation
by: Atsuki Osanai
Domain adaptation Recognition Representation learning Self supervised learning Text Semi-supervised Learning
Hyperbolic Vision Transformers: Combining Improvements in Metric Learning
by: Nakano Shuhei
Representation learning
Towards End-to-End Unified Scene Text Detection and Layout Analysis
by: Atsuki Osanai
Dataset Instance segmentation text layout analysis
Learning Local Displacements for Point Cloud Completion
by: Naoya Chiba
3D 3D reconstruction Point cloud Semantic segmentation
CAFE: Learning To Condense Dataset by Aligning Features
by: Yuma Ochi
Dataset
Forward Propagation, Backward Regression, and Pose Association for Hand Tracking in the Wild
by: Masanori YANO
Dataset Object tracking
Surface Representation for Point Clouds
by: Naoya Chiba
3D Point cloud Semantic segmentation
POCO: Point Convolution for Surface Reconstruction
by: 古川遼
3D reconstruction Point cloud
Diverse Plausible 360-Degree Image Outpainting for Efficient 3DCG Background Creation
by: Masanori YANO
Image completion Image outpainting Transformer
Dataset Distillation by Matching Training Trajectories
by: Yuma Ochi
Dataset
Camera Pose Estimation Using Implicit Distortion Models
by: shoji sonoyama
Pose estimation Camera calibration
Weakly Supervised Semantic Segmentation by Pixel-to-Prototype Contrast
by: Anonymous
Segmentation Semantic segmentation
Unsupervised Homography Estimation With Coplanarity-Aware GAN
by: Anonymous
GAN homography estimation
End-to-End Referring Video Object Segmentation With Multimodal Transformers
by: Ryuichi Nakahara
Segmentation Video Vision and language
Self-Supervised Arbitrary-Scale Point Clouds Upsampling via Implicit Neural Representation
by: Naoya Chiba
3D Point cloud Self supervised learning Super resolution
Masked Feature Prediction for Self-Supervised Visual Pre-Training
by: 志田　遥飛
Self supervised learning Video
Searching the Deployable Convolution Neural Networks for GPUs
by: Yuma Ochi
Neural architecture search(NAS)
Neural Point Light Fields
by: Masanori YANO
3D Point cloud NeRF
Vision Transformer With Deformable Attention
by: Kaikai Zhao
Object detection Recognition
Playable Environments: Video Manipulation in Space and Time
by: Akihiro FUJII
3D GAN NeRF
Temporally Efficient Vision Transformer for Video Instance Segmentation
by: Akihiro FUJII
Attetion Instance segmentation Video
Recurrent Glimpse-Based Decoder for Detection With Transformer
by: Anonymous
Object detection
Beyond Supervised vs. Unsupervised: Representative Benchmarking and Analysis of Image Representation Learning
by: kodai nakashima
Representation learning
TransRank: Self-Supervised Video Representation Learning via Ranking-Based Transformation Recognition
by: 志田　遥飛
Self supervised learning Video
Opening Up Open World Tracking
by: Masanori YANO
Dataset Object tracking
LAKe-Net: Topology-Aware Point Cloud Completion by Localizing Aligned Keypoints
by: Naoya Chiba
3D Point cloud Self supervised learning Super resolution
NeRFusion: Fusing Radiance Fields for Large-Scale Scene Reconstruction
by: 佐藤凜太郎
3D 3D reconstruction Neural radiance fields (NeRF)
Neural 3D Scene Reconstruction With the Manhattan-World Assumption
by: 佐藤凜太郎
3D 3D reconstruction Depth estimation Neural radiance fields (NeRF) Segmentation Semantic segmentation
Large-Scale Pre-Training for Person Re-Identification With Noisy Labels
by: Hirokatsu Kataoka
Person re-identification Self supervised learning
Tencent-MVSE: A Large-Scale Benchmark Dataset for Multi-Modal Video Similarity Evaluation
by: Hirokatsu Kataoka
Dataset Representation learning Video Vision and language
ZeroWaste Dataset: Towards Deformable Object Segmentation in Cluttered Scenes
by: Hirokatsu Kataoka
Dataset Instance segmentation
Proper Reuse of Image Classification Features Improves Object Detection
by: Anonymous
Object detection
The Auto Arborist Dataset: A Large-Scale Benchmark for Multiview Urban Forest Monitoring Under Domain Shift
by: Hirokatsu Kataoka
Dataset Object detection
DAIR-V2X: A Large-Scale Dataset for Vehicle-Infrastructure Cooperative 3D Object Detection
by: Hirokatsu Kataoka
3D object detection Dataset Point cloud
Knowledge Distillation As Efficient Pre-Training: Faster Convergence, Higher Data-Efficiency, and Better Transferability
by: 田所龍
Knowledge distillation Representation learning
RegNeRF: Regularizing Neural Radiance Fields for View Synthesis From Sparse Inputs
by: Hirokatsu Kataoka
Neural radiance fields (NeRF)
360MonoDepth: High-Resolution 360° Monocular Depth Estimation
by: Ryunosuke Isikawa
3D 3D object detection
DN-DETR: Accelerate DETR Training by Introducing Query DeNoising
by: Anonymous
Object detection
SHIFT: A Synthetic Driving Dataset for Continuous Multi-Task Domain Adaptation
by: Yuma Ochi
3D Dataset
Density-Preserving Deep Point Cloud Compression
by: Naoya Chiba
3D Point cloud Representation learning Self supervised learning
StyleMesh: Style Transfer for Indoor 3D Scene Reconstructions
by: Keiichi Sawada
3D 3D reconstruction
What Makes Transfer Learning Work for Medical Images: Feature Reuse & Other Factors
by: 奥西泰地
Domain adaptation Representation learning
ICON: Implicit Clothed Humans Obtained From Normals
by: Ryosuke Yamada
3D reconstruction
QueryDet: Cascaded Sparse Query for Accelerating High-Resolution Small Object Detection
by: Anonymous
Object detection
Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields
by: Hirokatsu Kataoka
Neural radiance fields (NeRF)
Killing Two Birds With One Stone: Efficient and Robust Training of Face Recognition CNNs by Partial FC
by: 鈴木共生
Robustness Face
TubeDETR: Spatio-Temporal Video Grounding With Transformers
by: Kazuki Omi
Multi modal Object detection Video Vision and language
Unsupervised Image-to-Image Translation With Generative Prior
by: Anonymous
GAN
SoftGroup for 3D Instance Segmentation on Point Clouds
by: Naoya Chiba
3D Object detection Point cloud Pose estimation Segmentation Semantic segmentation
Reflash Dropout in Image Super-Resolution
by: Masanori YANO
Super resolution Dropout
Iterative Deep Homography Estimation
by: Anonymous
homography estimation
Learning a Structured Latent Space for Unsupervised Point Cloud Completion
by: Naoya Chiba
3D Disentanglement Point cloud Representation learning Self supervised learning
Detecting Deepfakes With Self-Blended Images
by: 岡本大和
deepfake
Affine Medical Image Registration With Coarse-To-Fine Vision Transformer
by: cho
3D Attetion
Enhancing Adversarial Training With Second-Order Statistics of Weights
by: s ishikawa
Robustness
Everything at Once – Multi-Modal Fusion Transformer for Video Retrieval
by: Chihiro Nakatani（中谷 千洋）
Multi modal Video
SimMIM: A Simple Framework for Masked Image Modeling
by: Ryosuke Yamada
Representation learning Self supervised learning
Dual-AI: Dual-Path Actor Interaction Learning for Group Activity Recognition
by: Chihiro Nakatani（中谷 千洋）
Action recognition Recognition Video
X-Pool: Cross-Modal Language-Video Attention for Text-Video Retrieval
by: Ryuichi Nakahara
Video Vision and language
Video-Text Representation Learning via Differentiable Weak Temporal Alignment
by: Ryuichi Nakahara
Video Vision and language
Neural Template: Topology-Aware Reconstruction and Disentangled Generation of 3D Meshes
by: 古川遼
3D 3D reconstruction Disentanglement Mesh
Dynamic 3D Gaze From Afar: Deep Gaze Estimation From Temporal Eye-Head-Body Coordination
by: Anonymous
3D Action recognition Dataset Pose estimation
SimVQA: Exploring Simulated Environments for Visual Question Answering
by: QIUYUE
Dataset Vision and language
Contextual Debiasing for Visual Recognition With Causal Mechanisms
by: QIUYUE
Recognition
Aesthetic Text Logo Synthesis via Content-Aware Layout Inferring
by: Yoshi Truong
Dataset GAN Multi modal Vision and language
TableFormer: Table Structure Understanding With Transformers
by: 岡本大和（LINE Computer Vision Lab）
Object detection document analysis
Training Object Detectors From Scratch: An Empirical Study in the Era of Vision Transformer
by: Anonymous
Object detection
Scaling Up Your Kernels to 31x31: Revisiting Large Kernel Design in CNNs
by: cho
Attetion
Learning With Neighbor Consistency for Noisy Labels
by: take
Representation learning
MuKEA: Multimodal Knowledge Extraction and Accumulation for Knowledge-Based Visual Question Answering
by: Ryuichi Nakahara
Vision and language
Lepard: Learning Partial Point Cloud Matching in Rigid and Deformable Scenes
by: Naoya Chiba
3D Point cloud Pose estimation Self supervised learning
CLIPstyler: Image Style Transfer With a Single Text Condition
by: Takeru Endo
Multi modal Vision and language
UTC: A Unified Transformer With Inter-Task Contrastive Learning for Visual Dialog
by: Takeru Endo
Multi modal Vision and language
Confidence Propagation Cluster: Unleash Full Potential of Object Detectors
by: Anonymous
Object detection
Rethinking Deep Face Restoration
by: Masanori YANO
GAN Super resolution Face restoration
Spatial Commonsense Graph for Object Localisation in Partial Scenes
by: 綱島秀樹
3D 3D Object Localization GNN
BokehMe: When Neural Rendering Meets Classical Rendering
by: Anonymous
bokeh
Point-NeRF: Point-Based Neural Radiance Fields
by: 佐藤凜太郎
3D 3D reconstruction Depth estimation Neural radiance fields (NeRF) Point cloud
DeepFusion: Lidar-Camera Deep Fusion for Multi-Modal 3D Object Detection
by: Anonymous
3D object detection Point cloud
OmniFusion: 360 Monocular Depth Estimation via Geometry-Aware Fusion
by: Norikatsu Sumi
Depth estimation
Reconstructing Surfaces for Sparse Point Clouds With On-Surface Priors
by: Naoya Chiba
3D 3D reconstruction Point cloud Self supervised learning
Recurrent Dynamic Embedding for Video Object Segmentation
by: 上田　樹
Segmentation Video
CLIP-NeRF: Text-and-Image Driven Manipulation of Neural Radiance Fields
by: Norikatsu Sumi
Disentanglement Neural radiance fields (NeRF) Vision and language
Block-NeRF: Scalable Large Scene Neural View Synthesis
by: Hirokatsu Kataoka
Neural radiance fields (NeRF)
Rethinking Minimal Sufficient Representation in Contrastive Learning
by: take
Representation learning
Towards Driving-Oriented Metric for Lane Detection Models
by: Yuma Ochi
Dataset Domain adaptation
Surface Reconstruction From Point Clouds by Learning Predictive Context Priors
by: Naoya Chiba
3D Point cloud Self supervised learning
Object-Region Video Transformers
by: Shuhei M. Yoshida
Action recognition Object detection Video
Correlation Verification for Image Retrieval
by: Masanori YANO
Image retrieval
Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality
by: QIUYUE
Dataset Vision and language
M3L: Language-Based Video Editing via Multi-Modal Multi-Level Transformers
by: QIUYUE
Video Vision and language
IDEA-Net: Dynamic 3D Point Cloud Interpolation via Deep Embedding Alignment
by: Naoya Chiba
3D Disentanglement Point cloud Self supervised learning
Self-Supervised Models Are Continual Learners
by: Shida Haruhi
Self supervised learning
Global Tracking Transformers
by: Masanori YANO
Object tracking Transformer