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 : 17 Oct 2022
total number : 4
Intel Labs at Ego4D Challenge 2022: A Better Baseline for Audio-Visual Diarization
- 论文/Paper: http://arxiv.org/pdf/2210.07764
- 代码/Code: None
The Surprisingly Straightforward Scene Text Removal Method With Gated Attention and Region of Interest Generation: A Comprehensive Prominent Model Analysis
- 论文/Paper: http://arxiv.org/pdf/2210.07489
- 代码/Code: https://github.com/naver/garnet
Autoregressive Uncertainty Modeling for 3D Bounding Box Prediction
- 论文/Paper: http://arxiv.org/pdf/2210.07424
- 代码/Code: None
Task Grouping for Multilingual Text Recognition
- 论文/Paper: http://arxiv.org/pdf/2210.07423
- 代码/Code: None
CVPR2022
NeurIPS
Updated on : 17 Oct 2022
total number : 13
Learnable Polyphase Sampling for Shift Invariant and Equivariant Convolutional Networks
- 论文/Paper: http://arxiv.org/pdf/2210.08001
- 代码/Code: None
AVLEN: Audio-Visual-Language Embodied Navigation in 3D Environments
- 论文/Paper: http://arxiv.org/pdf/2210.07940
- 代码/Code: None
MOVE: Unsupervised Movable Object Segmentation and Detection
- 论文/Paper: http://arxiv.org/pdf/2210.07920
- 代码/Code: None
One Model to Edit Them All: Free-Form Text-Driven Image Manipulation with Semantic Modulations
- 论文/Paper: http://arxiv.org/pdf/2210.07883
- 代码/Code: page:https://github.com/KumapowerLIU/FFCLIP.
Model-Based Imitation Learning for Urban Driving
- 论文/Paper: http://arxiv.org/pdf/2210.07729
- 代码/Code: https://github.com/wayveai/mile.
Quo Vadis: Is Trajectory Forecasting the Key Towards Long-Term Multi-Object Tracking?
- 论文/Paper: http://arxiv.org/pdf/2210.07681
- 代码/Code: https://github.com/dendorferpatrick/QuoVadis.
DART: Articulated Hand Model with Diverse Accessories and Rich Textures
- 论文/Paper: http://arxiv.org/pdf/2210.07650
- 代码/Code: None
Mix and Reason: Reasoning over Semantic Topology with Data Mixing for Domain Generalization
- 论文/Paper: http://arxiv.org/pdf/2210.07571
- 代码/Code: None
TokenMixup: Efficient Attention-guided Token-level Data Augmentation for Transformers
- 论文/Paper: http://arxiv.org/pdf/2210.07562
- 代码/Code: https://github.com/mlvlab/tokenmixup
Weakly-Supervised Multi-Granularity Map Learning for Vision-and-Language Navigation
- 论文/Paper: http://arxiv.org/pdf/2210.07506
- 代码/Code: https://github.com/peihaochen/ws-mgmap
Learning Active Camera for Multi-Object Navigation
- 论文/Paper: http://arxiv.org/pdf/2210.07505
- 代码/Code: None
Evaluating Out-of-Distribution Performance on Document Image Classifiers
- 论文/Paper: http://arxiv.org/pdf/2210.07448
- 代码/Code: None
A Consistent and Differentiable Lp Canonical Calibration Error Estimator
- 论文/Paper: http://arxiv.org/pdf/2210.07810
- 代码/Code: https://github.com/tpopordanoska/ece-kde
