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
๐Ÿซฃ Text-Guided Adversarial Makeup ๐Ÿซฃ

๐Ÿ‘‰Novel facial privacy protection via adversarial latent codes. Makeup vs Face Recognition.

๐Ÿ˜ŽReview https://t.ly/pBCP
๐Ÿ˜ŽPaper arxiv.org/pdf/2306.10008.pdf
๐Ÿ˜ŽCode github.com/fahadshamshad/Clip2Protect
โค6๐Ÿ‘1๐Ÿ”ฅ1๐Ÿฅฐ1๐Ÿ’ฉ1
Media is too big
VIEW IN TELEGRAM
๐Ÿฆท Few-Shot Geometry-Aware Keypoints ๐Ÿฆท

๐Ÿ‘‰UBC (+Flawless AI) unveils the new SOTA in semantic keypoints localization. Suitable for faces, animals, cars, mouth, teeth & more

๐Ÿ˜ŽReview https://t.ly/-0qN
๐Ÿ˜ŽPaper arxiv.org/pdf/2303.17216.pdf
๐Ÿ˜ŽProject xingzhehe.github.io/FewShot3DKP/
๐Ÿคฏ10๐Ÿ‘4โค2โšก2๐Ÿ‘2๐Ÿคฉ2๐Ÿ”ฅ1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿš” Fooling Neural Forensic Classifiers ๐Ÿš”

๐Ÿ‘‰Adversarial faces able to fool the forensic classifiers, while remaining undetectable by humans

๐Ÿ˜ŽReview https://t.ly/33Cc
๐Ÿ˜ŽPaper arxiv.org/pdf/2306.13091.pdf
๐Ÿ˜ŽProject koushiksrivats.github.io/face_attribute_attack
๐Ÿ˜ŽCode github.com/koushiksrivats/face_attribute_attack
๐Ÿ˜ข6โค4๐Ÿ‘2๐Ÿ˜ฑ2๐Ÿพ2๐Ÿ‘1๐Ÿคฏ1๐Ÿ˜1
panohead_overview-min.gif
24.3 MB
๐Ÿฅ PanoHead: 3D Full-Head Synthesis ๐Ÿฅ

๐Ÿ‘‰#ByteDance (+UW-M) unveils PanoHead: 360โ—ฆ view-consistent portraits from a single-view image

๐Ÿ˜ŽReview https://t.ly/MrLNR
๐Ÿ˜ŽPaper arxiv.org/pdf/2303.13071.pdf
๐Ÿ˜ŽProject sizhean.github.io/panohead
๐Ÿ˜ŽCode github.com/sizhean/panohead
๐Ÿ”ฅ7โค4๐Ÿคฏ3๐Ÿ˜ฑ1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿ”ฎSAM-PT: Segment Anything+Tracking๐Ÿ”ฎ

๐Ÿ‘‰SAM-PT is the first method to utilize sparse point propagation for Video Object Segmentation (VOS).

๐Ÿ˜ŽReview https://t.ly/QLMG
๐Ÿ˜ŽPaper arxiv.org/pdf/2307.01197.pdf
๐Ÿ˜ŽProject www.vis.xyz/pub/sam-pt/
๐Ÿ˜ŽCode github.com/SysCV/sam-pt
๐Ÿ”ฅ14โค7๐Ÿคฏ3๐Ÿ‘1๐Ÿ˜ฑ1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿชฉ DISCO: Human Dance Generation ๐Ÿชฉ

๐Ÿ‘‰NTU (+ #Microsoft) unveils DISCO: a big step towards the Human Dance Generation.

๐Ÿ˜ŽReview https://t.ly/cNGX
๐Ÿ˜ŽPaper arxiv.org/pdf/2307.00040.pdf
๐Ÿ˜ŽProject disco-dance.github.io/
๐Ÿ˜ŽCode github.com/Wangt-CN/DisCo
๐Ÿ”ฅ13๐Ÿฅฐ4๐Ÿ˜2โšก1๐Ÿ‘1๐Ÿพ1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿ›ฃ๏ธ STAR.: 3D-tracking w/ attention paradigm ๐Ÿ›ฃ๏ธ

๐Ÿ‘‰#Mercedes STAR: e2e 3D object tracking that follows the tracking-by-attention paradigm

๐Ÿ˜ŽReview https://t.ly/JoGj
๐Ÿ˜ŽPaper arxiv.org/pdf/2306.17602.pdf
๐Ÿ˜ŽProject simondoll.github.io/publications/star_track
๐Ÿ‘14๐Ÿ”ฅ1๐Ÿฅฐ1๐Ÿ‘1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿก Text2Cinemagraphs: Cinemagraph from text ๐Ÿก

๐Ÿ‘‰CMU (+ #Snap) unveils a fully automated method for creating cinemagraphs from text descriptions

๐Ÿ˜ŽReview https://t.ly/BwZs6
๐Ÿ˜ŽPaper arxiv.org/pdf/2307.03190.pdf
๐Ÿ˜ŽProject text2cinemagraph.github.io/website
๐Ÿ˜ŽCode github.com/text2cinemagraph/text2cinemagraph
โค12๐Ÿคฏ3๐Ÿ˜ฑ1๐Ÿคฉ1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿ”ฅTest-Time Training on fire ๐Ÿ”ฅ

๐Ÿ‘‰Extending the TTT to the streaming setting. Suitable for Panoptic, Instance & Colorization.

๐Ÿ˜ŽReview https://t.ly/eZYA
๐Ÿ˜ŽPaper arxiv.org/pdf/2307.05014.pdf
๐Ÿ˜ŽProject https://video-ttt.github.io/
๐Ÿ˜ŽCode github.com/renwang435/video-ttt-release
๐Ÿ”ฅ10๐Ÿ‘3โšก1๐Ÿคฏ1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿƒ Deepfake via casual self-scan ๐Ÿƒ

๐Ÿ‘‰TAU presents a novel approach to reenact an ID using only a casual self-scan

๐Ÿ˜ŽReview https://t.ly/9T8Wi
๐Ÿ˜ŽPaper arxiv.org/pdf/2307.06307.pdf
๐Ÿ˜ŽProject arielazary.github.io/PGR
๐Ÿคฏ7๐Ÿ‘6โค5๐Ÿ”ฅ1๐Ÿ˜ฑ1
This media is not supported in your browser
VIEW IN TELEGRAM
๐ŸŽช Extreme Human Pose Estimation ๐ŸŽช

๐Ÿ‘‰RePoGen: novel synthetic data generator of extreme/realistic poses of humans

๐Ÿ˜ŽReview https://t.ly/ecBvM
๐Ÿ˜ŽPaper arxiv.org/pdf/2307.06737.pdf
๐Ÿ˜ŽProject mirapurkrabek.github.io/RePoGen-paper
๐Ÿ˜ŽCode github.com/MiraPurkrabek/RePoGen
๐Ÿ”ฅ12๐Ÿ‘2๐Ÿ‘1๐Ÿคฏ1๐Ÿ˜ฑ1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿ’ก DATID-3D: Text-to-3D Generation ๐Ÿ’ก

๐Ÿ‘‰ A novel domain adaptation method for 3D via text-to-image diffusion. ๐Ÿค—-Demo available!

๐Ÿ˜ŽReview https://t.ly/TCL-B
๐Ÿ˜ŽPaper arxiv.org/pdf/2211.16374.pdf
๐Ÿ˜ŽProject gwang-kim.github.io/datid_3d/
๐Ÿ˜ŽCode github.com/gwang-kim/DATID-3D
๐Ÿค— huggingface.co/spaces/gwang-kim/DATID-3D
๐Ÿ˜ŽColab colab.research.google.com/drive/1e9NSVB7x_hjz-nr4K0jO4rfTXILnNGtA?usp=sharing
๐Ÿคฏ5
This media is not supported in your browser
VIEW IN TELEGRAM
๐ŸงฏNeural Focal Modulation VAR๐Ÿงฏ

๐Ÿ‘‰A novel architecture for video recognition that models both local/global context

๐Ÿ˜ŽReview https://t.ly/rF_fk
๐Ÿ˜ŽPaper arxiv.org/pdf/2307.06947.pdf
๐Ÿ˜ŽProject talalwasim.github.io/Video-FocalNets
๐Ÿ˜ŽCode github.com/TalalWasim/Video-FocalNets
๐Ÿ”ฅ8โšก1๐Ÿ‘1๐Ÿคฉ1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿˆ Gen-AI as representation learner ๐Ÿˆ

๐Ÿ‘‰DreamTeacher: novel self-supervised feats. representation learning framework that utilizes gen-nets for pre-training downstream image backbones

๐Ÿ˜ŽReview https://t.ly/RL8iG
๐Ÿ˜ŽPaper arxiv.org/pdf/2307.07487.pdf
๐Ÿ˜ŽProject research.nvidia.com/labs/toronto-ai/DreamTeacher
๐Ÿ”ฅ9๐Ÿ‘2๐Ÿคฏ1
This media is not supported in your browser
VIEW IN TELEGRAM
โ˜” #SelfDriving? It's all about weather! โ˜”

๐Ÿ‘‰Novel self-supervised MDE method to handle adverse weather in real-world autonomous driving

๐Ÿ˜ŽReview https://t.ly/tcLQW
๐Ÿ˜ŽPaper arxiv.org/pdf/2307.08357.pdf
๐Ÿ˜ŽProject kieran514.github.io/Robust-Depth-Project/
โค7๐Ÿ‘3๐Ÿคฏ1๐Ÿ˜ฑ1
๐Ÿฆ™ Llama-2: the Open-Source "ChatGPT" ๐Ÿฆ™

๐Ÿ‘‰GenAI, #Meta unveils Llama-2: a collection of LLMs ranging in scale 7-70B params. Challenging with #chatgpt, but open.

๐Ÿ˜ŽReview https://t.ly/bLJgP
๐Ÿ˜ŽPaper https://t.ly/AOXru
๐Ÿ˜ŽProject https://ai.meta.com/llama
๐Ÿคฏ19โค2๐Ÿ”ฅ1๐Ÿ’ฉ1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿ‰ AltFreezing: new SOTA in detecting deepfake ๐Ÿ‰

๐Ÿ‘‰#Microsoft unveils AltFreezing: spatial/temporal artifacts in one model for more general face forgery detection

๐Ÿ˜ŽReview https://t.ly/mkIKX
๐Ÿ˜ŽPaper https://t.ly/z4KnJ
๐Ÿ˜ŽCode github.com/ZhendongWang6/AltFreezing
๐Ÿ˜ฑ6๐Ÿ‘5๐Ÿ˜4๐Ÿคฏ2๐Ÿฅฐ1
This media is not supported in your browser
VIEW IN TELEGRAM
๐ŸชŸMETA's Ultra-HD Data for #AR๐ŸชŸ

๐Ÿ‘‰Aria Digital Twin: egocentric dataset for detection/tracking, reconstruction/understanding, S2R learning, pose and more.

๐Ÿ˜ŽReview https://t.ly/MRPt1
๐Ÿ˜ŽPaper arxiv.org/pdf/2306.06362.pdf
๐Ÿ˜ŽProject www.projectaria.com/datasets/adt
๐Ÿ˜ŽCode github.com/facebookresearch/projectaria_tools
๐Ÿ”ฅ10๐Ÿ‘1
This media is not supported in your browser
VIEW IN TELEGRAM
๐Ÿ‘ฉโ€๐Ÿฆฐ Ultra-Realistic Neural Hair ๐Ÿ‘ฉโ€๐Ÿฆฐ

๐Ÿ‘‰A novel method to reconstruct the hair geometry at a strand level from monocular video or multi-view images

๐Ÿ˜ŽReview https://t.ly/6xZyp
๐Ÿ˜ŽPaper arxiv.org/pdf/2306.05872.pdf
๐Ÿ˜ŽProject samsunglabs.github.io/NeuralHaircut
๐Ÿ˜ŽCode github.com/SamsungLabs/NeuralHaircut
๐Ÿคฏ17๐Ÿคฉ5๐Ÿ˜5๐Ÿ‘2โšก1