论文速递2022.10.11！

最新成果demo展示：

代码：https://github.com/krisnarengga/yolov7-image-segmentation

配置好环境后运行指令：python segment/predict_counting.py --weights yolov7-seg.pt --source 120.mp4 --view-img --nosave --trk

最新论文整理

   ECCV2022

Updated on : 11 Oct 2022

total number : 10

SCAM! Transferring humans between images with Semantic Cross Attention Modulation

• 论文/Paper: http://arxiv.org/pdf/2210.04883

• 代码/Code: None

Using Whole Slide Image Representations from Self-Supervised Contrastive Learning for Melanoma Concordance Regression

• 论文/Paper: http://arxiv.org/pdf/2210.04803

• 代码/Code: None

BoundaryFace: A mining framework with noise label self-correction for Face Recognition

• 论文/Paper: http://arxiv.org/pdf/2210.04567

• 代码/Code: https://github.com/swjtu-3dvision/boundaryface

Self-Supervised 3D Human Pose Estimation in Static Video Via Neural Rendering

• 论文/Paper: http://arxiv.org/pdf/2210.04514

• 代码/Code: None

Students taught by multimodal teachers are superior action recognizers

• 论文/Paper: http://arxiv.org/pdf/2210.04331

• 代码/Code: None

Attention Diversification for Domain Generalization

• 论文/Paper: http://arxiv.org/pdf/2210.04206

• 代码/Code: https://github.com/hikvision-research/domaingeneralization

Fast-ParC: Position Aware Global Kernel for ConvNets and ViTs

• 论文/Paper: http://arxiv.org/pdf/2210.04020

• 代码/Code: None

FBNet: Feedback Network for Point Cloud Completion

• 论文/Paper: http://arxiv.org/pdf/2210.03974

• 代码/Code: https://github.com/hikvision-research/3dvision

Super-Resolution by Predicting Offsets: An Ultra-Efficient Super-Resolution Network for Rasterized Images

• 论文/Paper: http://arxiv.org/pdf/2210.04198

• 代码/Code: None

Strong Gravitational Lensing Parameter Estimation with Vision Transformer

• 论文/Paper: http://arxiv.org/pdf/2210.04143

• 代码/Code: https://github.com/kuanweih/strong_lensing_vit_resnet

    CVPR2022

   NeurIPS

Updated on : 11 Oct 2022

total number : 17

4D Unsupervised Object Discovery

• 论文/Paper: http://arxiv.org/pdf/2210.04801

• 代码/Code: https://github.com/robertwyq/lsmol

OGC: Unsupervised 3D Object Segmentation from Rigid Dynamics of Point Clouds

• 论文/Paper: http://arxiv.org/pdf/2210.04458

• 代码/Code: https://github.com/vlar-group/ogc

Semi-supervised Semantic Segmentation with Prototype-based Consistency Regularization

• 论文/Paper: http://arxiv.org/pdf/2210.04388

• 代码/Code: https://github.com/heimingx/semi_seg_proto

CAGroup3D: Class-Aware Grouping for 3D Object Detection on Point Clouds

• 论文/Paper: http://arxiv.org/pdf/2210.04264

• 代码/Code: https://github.com/haiyang-w/cagroup3d

Transformer-based Flood Scene Segmentation for Developing Countries

• 论文/Paper: http://arxiv.org/pdf/2210.04218

• 代码/Code: None

Let Images Give You More:Point Cloud Cross-Modal Training for Shape Analysis

• 论文/Paper: http://arxiv.org/pdf/2210.04208

• 代码/Code: https://github.com/zhanheshen/pointcmt

Coded Residual Transform for Generalizable Deep Metric Learning

• 论文/Paper: http://arxiv.org/pdf/2210.04180

• 代码/Code: None

Stimulative Training of Residual Networks: A Social Psychology Perspective of Loafing

• 论文/Paper: http://arxiv.org/pdf/2210.04153

• 代码/Code: https://github.com/Sunshine-Ye/NIPS22-ST

Robust Graph Structure Learning over Images via Multiple Statistical Tests

• 论文/Paper: http://arxiv.org/pdf/2210.03956

• 代码/Code: https://github.com/thomas-wyh/b-attention

Contact-aware Human Motion Forecasting

• 论文/Paper: http://arxiv.org/pdf/2210.03954

• 代码/Code: https://github.com/wei-mao-2019/contawaremotionpred

EgoTaskQA: Understanding Human Tasks in Egocentric Videos

• 论文/Paper: http://arxiv.org/pdf/2210.03929

• 代码/Code: None

ViewFool: Evaluating the Robustness of Visual Recognition to Adversarial Viewpoints

• 论文/Paper: http://arxiv.org/pdf/2210.03895

• 代码/Code: https://github.com/heathcliff-saku/viewfool_

FLamby: Datasets and Benchmarks for Cross-Silo Federated Learning in Realistic Healthcare Settings

• 论文/Paper: http://arxiv.org/pdf/2210.04620

• 代码/Code: https://github.com/owkin/flamby

Grow and Merge: A Unified Framework for Continuous Categories Discovery

• 论文/Paper: http://arxiv.org/pdf/2210.04174

• 代码/Code: None

Few-Shot Continual Active Learning by a Robot

• 论文/Paper: http://arxiv.org/pdf/2210.04137

• 代码/Code: None

Meta-DMoE: Adapting to Domain Shift by Meta-Distillation from Mixture-of-Experts

• 论文/Paper: http://arxiv.org/pdf/2210.03885

• 代码/Code: https://github.com/n3il666/Meta-DMoE.

In What Ways Are Deep Neural Networks Invariant and How Should We Measure This?

• 论文/Paper: http://arxiv.org/pdf/2210.03773

• 代码/Code: None