• 關(guān)注領(lǐng)域：

• AIGC
  

• 3D computer vision learning

• Fine-grained learning

• GNN

• 其他
  

• 聲明
  

• 論文較多，時(shí)間有限，本專欄無法做文章的講解，只挑選出符合PaperABC研究興趣和當(dāng)前熱點(diǎn)問題相關(guān)的論文，如果你的research topic和上述內(nèi)容有關(guān)，那本專欄可作為你的論文更新源或Paper reading list．

Paper list:

今日ArXiv共更新108篇

03.28共更新248篇

03.27共更新128篇

NeRF

HandNeRF: Neural Radiance Fields for Animatable Interacting Hands

https://arxiv.org/pdf/2303.13825.pdf

手勢(shì)仿真NeRF.

ABLE-NeRF: Attention-Based Rendering with Learnable Embeddings for Neural Radiance Field

https://arxiv.org/pdf/2303.13817.pdf

南洋理工S-Lab的工作．

GM-NeRF: Learning Generalizable Model-based Neural Radiance Fields from Multi-view Images

https://arxiv.org/pdf/2303.13777.pdf

UrbanGIRAFFE: Representing Urban Scenes as Compositional Generative Neural Feature Fields

https://arxiv.org/pdf/2303.14167.pdf

整體思想還是做組合物體生成，但是這篇文章聚焦urban scene.　手速真快?。?br>

AIGC

MindDiffuser: Controlled Image Reconstruction from Human Brain Activity with Semantic and Structural Diffusion

https://arxiv.org/pdf/2303.14139.pdf

基于腦信號(hào)來繪制圖像的文章，前段時(shí)間非?；穑?/p>

Fantasia3D: Disentangling Geometry and Appearance for High-quality Text-to-3D Content Creation

https://arxiv.org/pdf/2303.13873.pdf

解耦幾何和外觀的文本到3D的生成工作，這個(gè)領(lǐng)域近半年已經(jīng)卷成紅海了！

CompoNeRF: Text-guided Multi-object Compositional NeRF with Editable 3D Scene Layout

https://arxiv.org/pdf/2303.13843.pdf

又是一篇，組合多物體生成．來自于HKUST. 你們卷吧　我不玩了．

Conditional Image-to-Video Generation with Latent Flow Diffusion Models

https://arxiv.org/pdf/2303.13744.pdf

利用latent flow diffusion model 實(shí)現(xiàn)圖像到視頻的生成任務(wù)．

Anti-DreamBooth: Protecting users from personalized text-to-image synthesis

https://arxiv.org/pdf/2303.15433.pdf

一篇防止生成模型類似于DreamBooth，濫用圖像的工作．

Debiasing Scores and Prompts of 2D Diffusion for Robust Text-to-3D Generation

https://arxiv.org/pdf/2303.15413.pdf

真的是什么樣的idea都有，卷不動(dòng)這個(gè)領(lǐng)域．

Text-to-Image Diffusion Models are Zero-Shot Classifiers

https://arxiv.org/pdf/2303.15233.pdf

擴(kuò)散模型學(xué)習(xí)到的知識(shí)能用來分類嗎？很有啟發(fā)性的工作．

Make-It-3D: High-Fidelity 3D Creation from A Single Image with Diffusion Prior

https://arxiv.org/pdf/2303.14184.pdf

單張圖像生成3D物體，感覺做法和Magic3D很類似．

VLP

Best of Both Worlds: Multimodal Contrastive Learning with Tabular and Imaging Data

https://arxiv.org/pdf/2303.14080.pdf

表格數(shù)據(jù)和圖像數(shù)據(jù)的多模態(tài)預(yù)訓(xùn)練工作．

Accelerating Vision-Language Pretraining with Free Language Modeling

https://arxiv.org/pdf/2303.14038.pdf

VLP領(lǐng)域做加速的工作，來自騰訊團(tuán)隊(duì)．

Prompt Tuning based Adapter for Vision-Language Model Adaption

https://arxiv.org/pdf/2303.15234.pdf

利用prompt tuning來微調(diào)VLP大模型