清华&商汤提出了神经SDF！从多个照明条件下单视图纯阴影或RGB图像重建！论文/代码速递2022.11.29！

整理：AI算法与图像处理

CVPR2022论文和代码整理：https://github.com/DWCTOD/CVPR2022-Papers-with-Code-Demo

ECCV2022论文和代码整理：https://github.com/DWCTOD/ECCV2022-Papers-with-Code-Demo

最新成果demo展示：

标题：ShadowNeuS: Neural SDF Reconstruction by Shadow Ray Supervision

主页：https://gerwang.github.io/shadowneus/

摘要：

通过监督场景和多视图图像平面之间的相机光线，NeRF 为新视图合成任务重建神经场景表示。另一方面，光源和场景之间的阴影光线还有待考虑。因此，我们提出了一种新颖的阴影射线监督方案，可以优化沿射线的样本和射线位置。通过监督阴影光线，我们在多种光照条件下成功地从单视图纯阴影或 RGB 图像重建场景的神经 SDF。

最新论文整理

ECCV2022

Updated on : 29 Nov 2022

total number : 2

Boosting COVID-19 Severity Detection with Infection-aware Contrastive Mixup Classifcation

• 论文/Paper: http://arxiv.org/pdf/2211.14559
• 代码/Code: None

CMC v2: Towards More Accurate COVID-19 Detection with Discriminative Video Priors

• 论文/Paper: http://arxiv.org/pdf/2211.14557
• 代码/Code: None

CVPR2022

NeurIPS

Updated on : 29 Nov 2022

total number : 10

Hand-Object Interaction Image Generation

• 论文/Paper: http://arxiv.org/pdf/2211.15663
• 代码/Code: None

Investigating Prompt Engineering in Diffusion Models

• 论文/Paper: http://arxiv.org/pdf/2211.15462
• 代码/Code: None

Context-Adaptive Deep Neural Networks via Bridge-Mode Connectivity

• 论文/Paper: http://arxiv.org/pdf/2211.15436
• 代码/Code: None

Pitfalls of Conditional Batch Normalization for Contextual Multi-Modal Learning

• 论文/Paper: http://arxiv.org/pdf/2211.15071
• 代码/Code: None

Learning Dense Object Descriptors from Multiple Views for Low-shot Category Generalization

• 论文/Paper: http://arxiv.org/pdf/2211.15059
• 代码/Code: https://github.com/rehg-lab/dope_selfsup

Performance evaluation of deep segmentation models on Landsat-8 imagery

• 论文/Paper: http://arxiv.org/pdf/2211.14851
• 代码/Code: None

3D Reconstruction of Protein Complex Structures Using Synthesized Multi-View AFM Images

• 论文/Paper: http://arxiv.org/pdf/2211.14662
• 代码/Code: None

Unsupervised Wildfire Change Detection based on Contrastive Learning

• 论文/Paper: http://arxiv.org/pdf/2211.14654
• 代码/Code: None

DigGAN: Discriminator gradIent Gap Regularization for GAN Training with Limited Data

• 论文/Paper: http://arxiv.org/pdf/2211.14694
• 代码/Code: https://github.com/AilsaF/DigGAN

Where to Pay Attention in Sparse Training for Feature Selection?

• 论文/Paper: http://arxiv.org/pdf/2211.14627
• 代码/Code: None