AI with Papers - Artificial Intelligence & Deep Learning
15K subscribers
95 photos
235 videos
11 files
1.26K 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
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿฅ EVA3D: new SOTA in #3D humans ๐Ÿฅ

๐Ÿ‘‰EVA3D: new SOTA for unconditional NeRF-human generation from 2D only

๐Ÿ˜ŽReview https://bit.ly/3Th9qX7
๐Ÿ˜ŽCode github.com/hongfz16/EVA3D
๐Ÿ˜ŽPaper arxiv.org/pdf/2210.04888.pdf
๐Ÿ˜ŽProject hongfz16.github.io/projects/EVA3D.html
๐Ÿ”ฅ14๐Ÿ‘2
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿ f-DM: Diffusion Models by Apple ๐Ÿ

๐Ÿ‘‰Spectacular work by #Apple on DMs: HQ generation with better efficiency and semantic

๐Ÿ˜ŽReview https://bit.ly/3Tils2u
๐Ÿ˜ŽProject https://jiataogu.me/fdm/
๐Ÿ˜ŽPaper arxiv.org/pdf/2210.04955.pdf
โค10๐Ÿ˜ฑ2๐Ÿ‘1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿ…GENIE by #Nvidia -> Faster Generation๐Ÿ…

๐Ÿ‘‰Higher-Order Denoising Diffusion Solvers for faster and better synthesis

๐Ÿ˜ŽReview https://bit.ly/3CRjtwr
๐Ÿ˜ŽProject nv-tlabs.github.io/GENIE/
๐Ÿ˜ŽPaper arxiv.org/pdf/2210.05475.pdf
๐Ÿ˜ŽCode github.com/nv-tlabs/GENIE
๐Ÿ”ฅ10๐Ÿ‘4
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿฅฌ "Perception Test" by #DeepMind ๐Ÿฅฌ

๐Ÿ‘‰Huge dataset with obj & point tracks, temporal sounds, multiple & grounded vQA

๐Ÿ˜ŽReview https://bit.ly/3Vqh96Q
๐Ÿ˜ŽDataset github.com/deepmind/perception_test
๐Ÿ˜ŽProject www.deepmind.com/blog/measuring-perception-in-ai-models
๐Ÿ‘15๐Ÿ”ฅ4๐Ÿ˜ฑ3
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿ”ฅ Matterport 3D Semantics Dataset ๐Ÿ”ฅ

๐Ÿ‘‰#Meta opens HM3DSEM, the largest #3D real-world dataset with dense semantic

๐Ÿ˜ŽReview https://bit.ly/3yF4W4G
๐Ÿ˜ŽPaper arxiv.org/pdf/2210.05633.pdf
๐Ÿ˜ŽProject aihabitat.org/datasets/hm3d-semantics
๐Ÿ˜ŽData github.com/matterport/habitat-matterport-3dresearch
๐Ÿ‘13
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿฆ‘ Instant Map-free Relocalization ๐Ÿฆ‘

๐Ÿ‘‰#Niantic unveils a novel instant, metric scaled re-localization with one single photo

๐Ÿ˜ŽReview https://bit.ly/3S1Gdyh
๐Ÿ˜ŽPaper arxiv.org/pdf/2210.05494.pdf
๐Ÿ˜ŽProject research.nianticlabs.com/mapfree-reloc-benchmark
๐Ÿ˜ŽData research.nianticlabs.com/mapfree-reloc-benchmark/dataset
๐Ÿ”ฅ13๐Ÿ‘2
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿงฎ Novel DM for 3D Shapes by #Nvidia ๐Ÿงฎ

๐Ÿ‘‰Hierarchical Latent Point Diffusion Model (LION) for 3D shape generation

๐Ÿ˜ŽReview https://bit.ly/3yDhZ6I
๐Ÿ˜ŽPaper arxiv.org/pdf/2210.06978.pdf
๐Ÿ˜ŽProject https://nv-tlabs.github.io/LION/
๐Ÿ˜ŽCode(soon) github.com/nv-tlabs/LION
โค11๐Ÿ˜ฑ2๐Ÿ”ฅ1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿชฒ#6D estimation fully in the wild๐Ÿชฒ

๐Ÿ‘‰First ever self-supervised 6D pose estimation training in the wild

๐Ÿ˜ŽReview https://bit.ly/3yHdHuS
๐Ÿ˜ŽPaper arxiv.org/pdf/2210.07199.pdf
๐Ÿ˜ŽProject kywind.github.io/self-pose
๐Ÿ˜ŽCode (soon)
๐Ÿ‘15๐Ÿคฏ8๐Ÿ˜ฑ4
This media is not supported in your browser
VIEW IN TELEGRAM
โ›ฝ Stable Diffusion in #Blender โ›ฝ

๐Ÿ‘‰Render with SuperPowers: novel scene render via text prompt

๐Ÿ˜ŽReview https://bit.ly/3s1mEeN
๐Ÿ˜ŽCode github.com/benrugg/AI-Render
๐Ÿคฏ8๐Ÿ‘5โค2
This media is not supported in your browser
VIEW IN TELEGRAM
โšฝMarkerless Body-Object Interactionโšฝ

๐Ÿ‘‰Novel whole-bodies/objects interaction method from multi-view RGB-D data

๐Ÿ˜ŽReview https://bit.ly/3yO56GY
๐Ÿ˜ŽData intercap.is.tue.mpg.de/login.php
๐Ÿ˜ŽProject https://intercap.is.tue.mpg.de
๐Ÿ˜ŽCode github.com/YinghaoHuang91
๐Ÿ˜ŽPaper intercap.is.tue.mpg.de/media/upload/main.pdf
๐Ÿ”ฅ6๐Ÿ‘2๐Ÿพ1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿ”ฅ Dressing Avatars by #META ๐Ÿ”ฅ

๐Ÿ‘‰Novel deep photorealistic appearance method for physically-simulated clothing in #metaverse

๐Ÿ˜ŽReview https://bit.ly/3yRBW9Y
๐Ÿ˜ŽPaper arxiv.org/pdf/2206.15470.pdf
๐Ÿคฏ7๐Ÿ‘5๐Ÿพ2โค1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿช‚ Parallel NeRF for 6-DoF pose ๐Ÿช‚

๐Ÿ‘‰#Nvidia unveils a parallel NeRF for 6-DoF target pose estimation

๐Ÿ˜ŽReview https://bit.ly/3guWWwA
๐Ÿ˜ŽPaper arxiv.org/pdf/2210.10108.pdf
๐Ÿ˜ŽProject https://pnerfp.github.io/
๐Ÿ‘8๐Ÿ”ฅ3
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿฆ™LaMAR: Localization/Mapping for #AR๐Ÿฆ™

๐Ÿ‘‰A new benchmark for #AR in large and unconstrained scenes

๐Ÿ˜ŽReview https://bit.ly/3DjlnWU
๐Ÿ˜ŽPaper lamar.ethz.ch/files/LaMAR.pdf
๐Ÿ˜ŽProject https://lamar.ethz.ch/
๐Ÿ˜ŽCode github.com/microsoft/lamar-benchmark
๐Ÿ‘7๐Ÿ”ฅ4๐Ÿ’ฏ4
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿ”ฅNew SOTA in Panoptic Segmentation๐Ÿ”ฅ

๐Ÿ‘‰#Google (with Hinton๐Ÿคฏ) unveils Pix2Seq-D: novel generalist framework for panoptic segmentation

๐Ÿ˜ŽReview https://bit.ly/3DmpbGM
๐Ÿ˜ŽPaper arxiv.org/pdf/2210.06366.pdf
๐Ÿ”ฅ9๐Ÿ‘5๐Ÿคฏ3
This media is not supported in your browser
VIEW IN TELEGRAM
๐ŸŽจ UniColor: Unified Colorization ๐ŸŽจ

๐Ÿ‘‰The first unified framework for colorization via stroke, exemplar, text, and a mix of them

๐Ÿ˜ŽReview https://bit.ly/3gESR9y
๐Ÿ˜ŽPaper arxiv.org/pdf/2209.11223.pdf
๐Ÿ˜ŽProject luckyhzt.github.io/unicolor
๐Ÿ˜ŽCode (SOON)
๐Ÿคฏ18๐Ÿ”ฅ6๐Ÿ‘1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿคฏ Full-Body from head/hand signals ๐Ÿคฏ

๐Ÿ‘‰#Meta unveils AvatarPoser: first full-body pose method via userโ€™s head/hands

๐Ÿ˜ŽReview https://bit.ly/3gESR9y
๐Ÿ˜ŽPaper arxiv.org/pdf/2207.13784.pdf
๐Ÿ˜ŽCode github.com/eth-siplab/AvatarPoser
๐Ÿ‘9๐Ÿ‘3โค1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿค–JRBD: Egocentric Perception of Humans๐Ÿค–

๐Ÿ‘‰Stanford -> JRDB-Pose: Dataset with 600,000+ body pose annotations!

๐Ÿ˜ŽReview https://bit.ly/3gEZBE4
๐Ÿ˜ŽPaper arxiv.org/pdf/1910.11792.pdf
๐Ÿ˜ŽProject jrdb.erc.monash.edu/
๐Ÿ‘8๐Ÿ’ฏ4
This media is not supported in your browser
VIEW IN TELEGRAM
โ†•๏ธSOTA Action Detector @90+ FPS!โ†•๏ธ

๐Ÿ‘‰YOWO-plus: real-time method for spatio-temporal action detection. YOWO-Nano the fastest!

๐Ÿ˜ŽReview https://bit.ly/3TUdhcI
๐Ÿ˜ŽPaper arxiv.org/pdf/2210.11219.pdf
๐Ÿ˜ŽCode github.com/yjh0410/PyTorch_YOWO
๐Ÿ‘13๐Ÿฅฐ2
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿ›Ž๏ธ๐Ÿ›Ž๏ธAutoregressive NeRF-Avatar๐Ÿ›Ž๏ธ๐Ÿ›Ž๏ธ

๐Ÿ‘‰AutoAvatar by #Meta: autoregressive method for modeling dynamically deforming human bodies from raw scans

๐Ÿ˜ŽReview https://bit.ly/3W0oTgo
๐Ÿ˜ŽPaper arxiv.org/pdf/2203.13817.pdf
๐Ÿ˜ŽProject zqbai-jeremy.github.io/autoavatar
๐Ÿ˜ŽCode github.com/facebookresearch/AutoAvatar
๐Ÿ‘11๐Ÿ”ฅ2๐Ÿพ1