ECCV2022 &CVPR2022论文速递2022.7.6！

2022-12-11 11:05:23 浏览数 (1)

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：

ECCV2022 | FlowFormer: 基于transformer架构的光流

标题：FlowFormer: A Transformer Architecture for Optical Flow 论文：https://arxiv.org/abs/2203.16194

主页：https://drinkingcoder.github.io/publication/flowformer/

摘要：本文介绍了光流transformer（FlowFormer），一种基于transformer的神经网络架构，用于学习光流。FlowFormer 对从图像对构建的 4D 成本量进行标记，将cost token 编码到具有新颖潜在空间中的交替组transformer(AGT) 层的成本存储器中，并通过具有动态位置成本查询的循环变换器解码器对成本存储器进行解码 . 在 Sintel 基准测试中，FlowFormer 实现了 1.178 的平均端点误差 (AEPE)，与公布的最佳结果 (1.388) 相比，误差减少了 15.1%。此外，FlowFormer 还实现了强大的泛化性能。在没有接受 Sintel 训练的情况下，FlowFormer 在 Sintel 训练集干净通行证上达到 1.00 AEPE，比公布的最佳结果 (1.29) 高出 22.4%。

最新论文整理：

Updated on : 6 Jul 2022

total number : 9

Detecting and Recovering Sequential DeepFake Manipulation

  • 论文/Paper: http://arxiv.org/abs/2207.02204
  • 代码/Code: https://github.com/rshaojimmy/seqdeepfake

Accelerating Score-based Generative Models with Preconditioned Diffusion Sampling

  • 论文/Paper: http://arxiv.org/abs/2207.02196
  • 代码/Code: https://github.com/fudan-zvg/pds

Improving Covariance Conditioning of the SVD Meta-layer by Orthogonality

  • 论文/Paper: http://arxiv.org/abs/2207.02119
  • 代码/Code: https://github.com/kingjamessong/orthoimprovecond

RepMix: Representation Mixing for Robust Attribution of Synthesized Images

  • 论文/Paper: http://arxiv.org/abs/2207.02063
  • 代码/Code: https://github.com/tubui/image_attribution

AvatarCap: Animatable Avatar Conditioned Monocular Human Volumetric Capture

  • 论文/Paper: http://arxiv.org/abs/2207.02031
  • 代码/Code: https://github.com/lizhe00/AvatarCap.

DeepPS2: Revisiting Photometric Stereo Using Two Differently Illuminated Images

  • 论文/Paper: http://arxiv.org/abs/2207.02025
  • 代码/Code: None

Learning Local Implicit Fourier Representation for Image Warping

  • 论文/Paper: http://arxiv.org/abs/2207.01831
  • 代码/Code: https://github.com/jaewon-lee-b/ltew

SESS: Saliency Enhancing with Scaling and Sliding

  • 论文/Paper: http://arxiv.org/abs/2207.01769
  • 代码/Code: https://github.com/neouyghur/sess

TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and Texts

  • 论文/Paper: http://arxiv.org/abs/2207.01696
  • 代码/Code: None

Updated on : 6 Jul 2022

total number : 4

Latents2Segments: Disentangling the Latent Space of Generative Models for Semantic Segmentation of Face Images

  • 论文/Paper: http://arxiv.org/abs/2207.01871
  • 代码/Code: None

GP22: A Car Styling Dataset for Automotive Designers

  • 论文/Paper: http://arxiv.org/abs/2207.01760
  • 代码/Code: None

How Much More Data Do I Need? Estimating Requirements for Downstream Tasks

  • 论文/Paper: http://arxiv.org/abs/2207.01725
  • 代码/Code: None

Egocentric Video-Language Pretraining @ Ego4D Challenge 2022

  • 论文/Paper: http://arxiv.org/abs/2207.01622
  • 代码/Code: https://github.com/showlab/egovlp
编程算法 http https 网络安全 github

0 人点赞