AI with Papers - Artificial Intelligence & Deep Learning
15K subscribers
96 photos
238 videos
11 files
1.27K links
All the AI with papers. Every day fresh updates on Deep Learning, Machine Learning, and Computer Vision (with Papers).

Curated by Alessandro Ferrari | https://www.linkedin.com/in/visionarynet/
Download Telegram
πŸ“Hyper-Dense Landmarks at 150FPSπŸ“

πŸ‘‰#Microsoft unveils the SOTA in dense landmarking + #3D reconstruction. MAGIC.

𝐇𝐒𝐠𝐑π₯𝐒𝐠𝐑𝐭𝐬:
βœ…Accurate 10Γ— as many landmarks as usual
βœ…Synthetic data, perfect annotations
βœ…NO appearance, light, diff-rendering
βœ…#3D @150+FPS with a single CPU thread
βœ…SOTA in monocular 3D reconstruction

More: https://bit.ly/37pQS40
πŸ‘6πŸ”₯4🀯1
This media is not supported in your browser
VIEW IN TELEGRAM
πŸͺ°NUWA-Infinity is out!πŸͺ°

πŸ‘‰βˆž generation by #Microsoft: arbitrarily-sized HD images and long videos 🀯

𝐇𝐒𝐠𝐑π₯𝐒𝐠𝐑𝐭𝐬:
βœ…Unconditional Image Gen.
βœ…Text-to-Image/Text-to-Clip
βœ…Animation / Out-painting
βœ…Hi-res, arbitrary long clip
βœ…NCP for patches caching

More: https://bit.ly/3zmBf9f
πŸ”₯7πŸ‘2❀1πŸ‘1🀯1
This media is not supported in your browser
VIEW IN TELEGRAM
🧰 FGT: flow-guided inpainting 🧰

πŸ‘‰#Microsoft (+USTC) unveils FGT: flow-guided ViT for video inpainting 🀯

𝐇𝐒𝐠𝐑π₯𝐒𝐠𝐑𝐭𝐬:
βœ…OF into transformer for attention++
βœ…Flow completion net w/ local feats.
βœ…Dual perspective spatial MHSA
βœ…Local attention with global content

More: https://bit.ly/3pk5J5S
❀11πŸ‘5
This media is not supported in your browser
VIEW IN TELEGRAM
😈 Synthetic Expression-Wrinkles 😈

πŸ‘‰#Microsoft unveils a novel approach that produces realistic wrinkles across humans

😎Review https://bit.ly/3zWZLOd
😎Paper arxiv.org/pdf/2210.03529.pdf
😎Project microsoft.github.io/DynamicWrinkles
πŸ”₯7🀯4πŸ‘2😱1
This media is not supported in your browser
VIEW IN TELEGRAM
πŸͺ΄ Rodin: 3D Avatars Using Diffusion πŸͺ΄

πŸ‘‰#Microsoft unveils a novel #3D diffusion for digital avatars as NeRF

😎Review https://bit.ly/3jcxeOX
😎Project 3d-avatar-diffusion.microsoft.com
😎Paper arxiv.org/pdf/2212.06135.pdf
❀9🀯4πŸ‘2😱1
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ—£οΈ MemFace: Generative Talking Face πŸ—£οΈ

πŸ‘‰#Microsoft (+SJTU) unveils MemFace: the new SOTA in talking faces generation

😎Review https://bit.ly/3k8TjhZ
😎Paper arxiv.org/pdf/2212.05005v2.pdf
😎Project memoryface.github.io/
🀯12🀩3πŸ‘1
This media is not supported in your browser
VIEW IN TELEGRAM
πŸͺ© DISCO: Human Dance Generation πŸͺ©

πŸ‘‰NTU (+ #Microsoft) unveils DISCO: a big step towards the Human Dance Generation.

😎Review https://t.ly/cNGX
😎Paper arxiv.org/pdf/2307.00040.pdf
😎Project disco-dance.github.io/
😎Code github.com/Wangt-CN/DisCo
πŸ”₯13πŸ₯°4😍2⚑1πŸ‘1🍾1
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ‰ AltFreezing: new SOTA in detecting deepfake πŸ‰

πŸ‘‰#Microsoft unveils AltFreezing: spatial/temporal artifacts in one model for more general face forgery detection

😎Review https://t.ly/mkIKX
😎Paper https://t.ly/z4KnJ
😎Code github.com/ZhendongWang6/AltFreezing
😱6πŸ‘5😍4🀯2πŸ₯°1
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ„ Video Understanding with GPT-4V(ision) πŸ„

πŸ‘‰ #Microsoft unveils MM-Vid, the most advanced video understanding framework (w/ #chatgpt4). Impressive results on long-form videos & intricate tasks such as audio description & multimodal high-level comprehension

😎Review https://t.ly/RISMm
😎Paper arxiv.org/pdf/2310.19773.pdf
😎Project https://multimodal-vid.github.io
🀯22πŸ‘9πŸ”₯2πŸ‘1😱1
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ”₯Florence-2: unified Computer VisionπŸ”₯

πŸ‘‰#Microsoft announces Florence-2: novel foundation model with unified, prompt-based, representation for a large variety of #computervision & vision-language task. One backbone -> multiple tasks!

πŸ‘‰Review https://t.ly/pOins
πŸ‘‰Paper arxiv.org/pdf/2311.06242.pdf
πŸ‘‰Project www.microsoft.com/en-us/research/project/projectflorence/
😱9❀5πŸ”₯3πŸ‘1πŸ‘1🍾1
This media is not supported in your browser
VIEW IN TELEGRAM
🩰 Dressed Humans in the wild 🩰

πŸ‘‰ETH (+ #Microsoft ) ReLoo: novel 3D-HQ reconstruction of humans dressed in loose garments from mono in-the-wild clips. No prior assumptions about the garments. Source Code announced, coming πŸ’™

πŸ‘‰Review https://t.ly/evgmN
πŸ‘‰Paper arxiv.org/pdf/2409.15269
πŸ‘‰Project moygcc.github.io/ReLoo/
πŸ‘‰Code github.com/eth-ait/ReLoo
🀯9❀2πŸ‘1πŸ”₯1
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ”₯BitNet: code of 1-bit LLM releasedπŸ”₯

πŸ‘‰BitNet by #Microsoft, announced in late 2023, is a 1-bit Transformer architecture designed for LLMs. BitLinear as a drop-in replacement of the nn.Linear layer in order to train 1-bit weights from scratch. Source Code just released πŸ’™

πŸ‘‰Review https://t.ly/3G2LA
πŸ‘‰Paper arxiv.org/pdf/2310.11453
πŸ‘‰Code https://lnkd.in/duPADJVb
πŸ”₯21❀5🀯2πŸ‘1πŸ₯°1
This media is not supported in your browser
VIEW IN TELEGRAM
🧿 Look Ma, no markers 🧿

πŸ‘‰#Microsoft unveils the first technique for marker-free, HQ reconstruction of COMPLETE human body, including eyes and tongue, without requiring any calibration, manual intervention or custom hardware. Impressive results! Repo for training & Dataset releasedπŸ’™

πŸ‘‰Review https://t.ly/5fN0g
πŸ‘‰Paper arxiv.org/pdf/2410.11520
πŸ‘‰Project microsoft.github.io/SynthMoCap/
πŸ‘‰Repo github.com/microsoft/SynthMoCap
🀯16πŸ‘10πŸ”₯3😱3❀1πŸ‘1
This media is not supported in your browser
VIEW IN TELEGRAM
🌈DAViD: Synthetic Depth-Normal-Segmentation🌈

πŸ‘‰#Microsoft's DAViD: 100% synthetic dataset/models for human Depth, Normals & Segmentation. Dataset available, models & runtime under MITπŸ’™

πŸ‘‰Review https://t.ly/-SlO_
πŸ‘‰Paper https://lnkd.in/eCmMXpTg
πŸ‘‰Project https://lnkd.in/eurCSWkm
πŸ‘‰Repo https://lnkd.in/e7PWFgP2
πŸ‘6❀4πŸ”₯2🀩1