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:CLIPascene: Scene Sketching with Different Types and Levels of Abstraction
主页:https://clipascene.github.io/CLIPascene/
论文:https://arxiv.org/abs/2211.17256
代码:未开源
摘要:
在本文中,我们提出了一种使用不同类型和多层次抽象将给定场景图像转换为草图的方法。我们区分两种类型的抽象。第一个考虑草图的保真度,将其表示从更精确的输入描绘变为更宽松的描绘。第二个是由草图的视觉简单性定义的,从详细的描绘转变为稀疏的草图。使用明确分解为两个抽象轴——每个抽象轴有多个层次——为用户提供了额外的控制,可以根据他们的个人目标和偏好选择所需的草图。为了以给定的保真度和简化程度形成草图,我们训练了两个 MLP 网络。第一个网络学习所需的笔画位置,而第二个网络学习在不损害其可识别性和语义的情况下逐渐从草图中删除笔画。我们的方法能够生成复杂场景的草图,包括具有复杂背景(例如,自然和城市环境)和主题(例如,动物和人)的场景,同时根据保真度和简单性描绘输入场景的渐进抽象。
最新论文整理
ECCV2022
Updated on : 7 Dec 2022
total number : 3
Union-set Multi-source Model Adaptation for Semantic Segmentation
- 论文/Paper: http://arxiv.org/pdf/2212.02785
- 代码/Code: https://github.com/lzy7976/union-set-model-adaptation.
Pixel2ISDF: Implicit Signed Distance Fields based Human Body Model from Multi-view and Multi-pose Images
- 论文/Paper: http://arxiv.org/pdf/2212.02765
- 代码/Code: None
QFT: Post-training quantization via fast joint finetuning of all degrees of freedom
- 论文/Paper: http://arxiv.org/pdf/2212.02634
- 代码/Code: None
CVPR2022
NeurIPS
Updated on : 7 Dec 2022
total number : 4
Land Use Prediction using Electro-Optical to SAR Few-Shot Transfer Learning
- 论文/Paper: http://arxiv.org/pdf/2212.03084
- 代码/Code: None
Super-resolution Probabilistic Rain Prediction from Satellite Data Using 3D U-Nets and EarthFormers
- 论文/Paper: http://arxiv.org/pdf/2212.02998
- 代码/Code: https://github.com/bugsuse/weather4cast-2022-stage2
Simple Baseline for Weather Forecasting Using Spatiotemporal Context Aggregation Network
- 论文/Paper: http://arxiv.org/pdf/2212.02952
- 代码/Code: None
Continual learning on deployment pipelines for Machine Learning Systems
- 论文/Paper: http://arxiv.org/pdf/2212.02659
- 代码/Code: None