ECCV2022 | 超越 SPADE，SCAM语义生成图像能应对更具挑战性的任务! 论文/代码速递2022.10.14！

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：SCAM! Transferring humans between images with Semantic Cross Attention Modulation

主页：https://imagine.enpc.fr/~dufourn/publications/scam.html

代码：https://github.com/nicolas-dufour/SCAM

论文：https://arxiv.org/pdf/2210.04883v1.pdf

最近的大量工作以语义条件下的图像生成为目标。大多数这类方法只关注较窄的姿势转移任务，而忽略了更具挑战性的对象转移任务，即不仅转移姿势，还转移外观和背景。在这项工作中，我们引入了SCAM（Semantic Cross Attention Modulation，语义交叉注意调制），这是一个系统，它对图像的每个语义区域（包括前景和背景）中丰富多样的信息进行编码，从而实现了以细节为重点的精确生成。这是由Semantic Attention Transformer Encoder实现的，该编码器为每个语义区域提取多个潜在向量，以及通过使用语义交叉注意调制来利用这些潜在向量的相应生成器。它仅使用重建设置进行训练，而受试者在测试时进行转移。我们的分析表明，我们提出的架构在编码每个语义区域的外观多样性方面是成功的。iDesigner和CelebAMask HD数据集上的大量实验表明，SCAM优于SEAN和SPADE；此外，它还开创了学科转移的新境界。

最新论文整理

ECCV2022

Updated on : 14 Oct 2022

total number : 1

Improving the Reliability for Confidence Estimation

• 论文/Paper: http://arxiv.org/pdf/2210.06776
• 代码/Code: None

CVPR2022

NeurIPS

Updated on : 14 Oct 2022

total number : 17

OpenOOD: Benchmarking Generalized Out-of-Distribution Detection

• 论文/Paper: http://arxiv.org/pdf/2210.07242
• 代码/Code: https://github.com/Jingkang50/OpenOOD

HSurf-Net: Normal Estimation for 3D Point Clouds by Learning Hyper Surfaces

• 论文/Paper: http://arxiv.org/pdf/2210.07158
• 代码/Code: None

RTFormer: Efficient Design for Real-Time Semantic Segmentation with Transformer

• 论文/Paper: http://arxiv.org/pdf/2210.07124
• 代码/Code: PaddleSeg:https://github.com/PaddlePaddle/PaddleSeg.

LION: Latent Point Diffusion Models for 3D Shape Generation

• 论文/Paper: http://arxiv.org/pdf/2210.06978
• 代码/Code: None

SageMix: Saliency-Guided Mixup for Point Clouds

• 论文/Paper: http://arxiv.org/pdf/2210.06944
• 代码/Code: https://github.com/mlvlab/SageMix

Feature-Proxy Transformer for Few-Shot Segmentation

• 论文/Paper: http://arxiv.org/pdf/2210.06908
• 代码/Code: https://github.com/Jarvis73/FPTrans

Adv-Attribute: Inconspicuous and Transferable Adversarial Attack on Face Recognition

• 论文/Paper: http://arxiv.org/pdf/2210.06871
• 代码/Code: None

Scalable Neural Video Representations with Learnable Positional Features

• 论文/Paper: http://arxiv.org/pdf/2210.06823
• 代码/Code: None

ALIFE: Adaptive Logit Regularizer and Feature Replay for Incremental Semantic Segmentation

• 论文/Paper: http://arxiv.org/pdf/2210.06816
• 代码/Code: None

Intermediate Prototype Mining Transformer for Few-Shot Semantic Segmentation

• 论文/Paper: http://arxiv.org/pdf/2210.06780
• 代码/Code: https://github.com/LIUYUANWEI98/IPMT

Q-ViT: Accurate and Fully Quantized Low-bit Vision Transformer

• 论文/Paper: http://arxiv.org/pdf/2210.06707
• 代码/Code: onhttps://github.com/YanjingLi0202/Q-ViT

Structural Pruning via Latency-Saliency Knapsack

• 论文/Paper: http://arxiv.org/pdf/2210.06659
• 代码/Code: None

S4ND: Modeling Images and Videos as Multidimensional Signals Using State Spaces

• 论文/Paper: http://arxiv.org/pdf/2210.06583
• 代码/Code: None

Task-Free Continual Learning via Online Discrepancy Distance Learning

• 论文/Paper: http://arxiv.org/pdf/2210.06579
• 代码/Code: None

Flare7K: A Phenomenological Nighttime Flare Removal Dataset

• 论文/Paper: http://arxiv.org/pdf/2210.06570
• 代码/Code: None

PDEBENCH: An Extensive Benchmark for Scientific Machine Learning

• 论文/Paper: http://arxiv.org/pdf/2210.07182
• 代码/Code: https://github.com/pdebench/PDEBench.

Brain Network Transformer

• 论文/Paper: http://arxiv.org/pdf/2210.06681
• 代码/Code: https://github.com/Wayfear/BrainNetworkTransformer.