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:
主页: https://omriavrahami.com/blended-diffusion-page/ 代码: https://github.com/omriav/blended-diffusion 摘要:
在本文中,我们介绍了第一个基于自然语言描述和 ROI 掩码在通用自然图像中执行局部(基于区域)编辑的解决方案。我们通过利用和组合预训练的语言图像模型 (CLIP) 来实现我们的目标,将编辑转向用户提供的文本提示,并使用去噪扩散概率模型 (DDPM) 来生成看起来自然的结果。展示了几个文本驱动的编辑应用程序,包括向图像添加新对象、删除/替换/更改现有对象、背景替换和图像外推。
最新论文整理
ECCV2022
Updated on : 16 Sep 2022
total number : 6
Hydra Attention: Efficient Attention with Many Heads
- 论文/Paper: http://arxiv.org/pdf/2209.07484
- 代码/Code: None
Self-distilled Feature Aggregation for Self-supervised Monocular Depth Estimation
- 论文/Paper: http://arxiv.org/pdf/2209.07088
- 代码/Code: https://github.com/ZM-Zhou/SDFA-Net_pytorch
MIPI 2022 Challenge on RGB ToF Depth Completion: Dataset and Report
- 论文/Paper: http://arxiv.org/pdf/2209.07057
- 代码/Code: https://github.com/mipi-challenge/MIPI2022.
MIPI 2022 Challenge on Quad-Bayer Re-mosaic: Dataset and Report
- 论文/Paper: http://arxiv.org/pdf/2209.07060
- 代码/Code: https://github.com/mipi-challenge/MIPI2022.
MIPI 2022 Challenge on Under-Display Camera Image Restoration: Methods and Results
- 论文/Paper: http://arxiv.org/pdf/2209.07052
- 代码/Code: https://github.com/mipi-challenge/MIPI2022.
Lossy Image Compression with Conditional Diffusion Models
- 论文/Paper: http://arxiv.org/pdf/2209.06950
- 代码/Code: None