🌵POCO: 3D HPS + Confidence🌵

👉 Novel framework for HPS: #3D human body + confidence in a single feed-forward pass

😎Review https://t.ly/cDePe
😎Paper arxiv.org/pdf/2308.12965.pdf
😎Project https://poco.is.tue.mpg.de

🌆 NeO360: NeRF for Sparse Outdoor 🌆

👉#Toyota (+GIT) unveils NeO360: 360◦ outdoor scenes from a single or a few posed RGB images

😎Review https://t.ly/JDJZg
😎Paper arxiv.org/pdf/2308.12967.pdf
😎Project zubair-irshad.github.io/projects/neo360.html

🥕 Scenimefy: I-2-I for anime 🥕

👉S-Lab unveils a novel semi-supervised I-2-I translation framework + HD dataset for anime

😎Review https://t.ly/IsdEG
😎Paper arxiv.org/pdf/2308.12968.pdf
😎Code https://github.com/Yuxinn-J/Scenimefy
😎Project https://yuxinn-j.github.io/projects/Scenimefy.html

🐨 Watch Your Steps: Editing by Text 🐨

👉The novel SOTA in image & scene (text) editing via denoising diffusion models

😎Review https://t.ly/fv9wn
😎Paper arxiv.org/pdf/2308.08947.pdf
😎Project ashmrz.github.io/WatchYourSteps

💡 Relighting NeRF 💡

👉Neural implicit radiance representation for free viewpoint relighting of an object lit by a moving point light

😎Review https://t.ly/J-3_L
😎Project nrhints.github.io
😎Code github.com/iamNCJ/NRHints
😎Paper nrhints.github.io/pdfs/nrhints-sig23.pdf

🪶 ReST: Multi-Camera MOT 🪶

👉Novel reconfigurable two-steps graph model for multi-camera multi object video tracking (MC-MOT)

😎Review https://t.ly/3C5tb
😎Paper arxiv.org/pdf/2308.13229.pdf
😎Code github.com/chengche6230/ReST

🌲MagicEdit: Magic Video Edit🌲

👉MagicEdit: explicit disentangling content, structure & motion for Hi-Fi and temporally coherent video editing

😎Report https://t.ly/tREX4
😎Paper arxiv.org/pdf/2308.14749.pdf
😎Project magic-edit.github.io
😎Code github.com/magic-research/magic-edit

✂️ VideoCutLER: Simple UVIS ✂️

👉VideoCutLER is a simple unsupervised video instance segmentation (UVIS) method without relying on optical flows

😎Review https://t.ly/PBBjG
😎Paper arxiv.org/pdf/2308.14710.pdf
😎Project people.eecs.berkeley.edu/~xdwang/projects/CutLER
😎Code github.com/facebookresearch/CutLER/tree/main/videocutler

🐦 3D Pigeons Pose & Tracking 🐦

👉 3D-MuPPET: estimate and track 3D poses of pigeons with multiple-views

😎Review https://t.ly/jfAJJ
😎Paper arxiv.org/pdf/2308.15316.pdf
😎Code github.com/alexhang212/3D-MuPPET/

🎍RoboTAP: Dense Tracking for Few-Shot Imitation🎍

👉RoboTAP: novel dense tracking representation for robotic arm

😎Review https://t.ly/MCO_V
😎Paper arxiv.org/pdf/2308.15975.pdf
😎Project https://robotap.github.io/
😎Code github.com/deepmind/tapnet

⛺FACET: Fairness in Computer Vision⛺

👉#META AI opens a large, publicly available dataset for classification, detection & segmentation. Potential performance disparities & challenges across sensitive demographic attributes

😎Review https://t.ly/mKn-t
😎Paper arxiv.org/pdf/2309.00035.pdf
😎Dataset https://facet.iss.onetademolab.com/

♊️ Doppelgangers in Structures ♊️

👉A novel learning-based approach for visual disambiguation: distinguishing illusory matches to produce correct, disambiguated #3D reconstructions

😎Review https://t.ly/9yLot
😎Paper arxiv.org/pdf/2309.02420.pdf
😎Code github.com/RuojinCai/Doppelgangers
😎Project doppelgangers-3d.github.io/

🍃 Tracking Anything with Decoupled VOS 🍃

👉A novel VOS approach that extends SAM for open-world video segmentation with no user input required

😎Review https://t.ly/xeobR
😎Paper arxiv.org/pdf/2309.03903.pdf
😎Project hkchengrex.com/Tracking-Anything-with-DEVA
😎Code github.com/hkchengrex/Tracking-Anything-with-DEVA
😎Colab https://colab.research.google.com/drive/1OsyNVoV_7ETD1zIE8UWxL3NXxu12m_YZ

🪷 Diffusive Consistent Video Editing 🪷

👉 Weizmann Institute of Science unveils TokenFlow, a novel text-to-image diffusion model for text-driven video editing

😎Review https://t.ly/ru8km
😎Paper arxiv.org/pdf/2307.10373.pdf
😎Project diffusion-tokenflow.github.io
😎Code github.com/omerbt/TokenFlow

🔥🔥 #META's DINOv2 is now commercial! 🔥🔥

👉Universal features for image classification, instance retrieval, video understanding, depth & semantic segmentation. Now suitable for commercial.

😎Review https://t.ly/LNrGy
😎Paper arxiv.org/pdf/2304.07193.pdf
😎Code github.com/facebookresearch/dinov2
😎Demo dinov2.metademolab.com/

🧄FreeMan: towards #3D Humans 🧄

👉FreeMan: the first large-scale, real-world, multi-view dataset for #3D human pose estimation. 11M frames!

😎Review https://t.ly/ICxpA
😎Paper arxiv.org/pdf/2309.05073.pdf
😎Project wangjiongw.github.io/freeman

🦊 MagiCapture: HD Multi-Concept Portrait 🦊

👉KAIST unveils MagiCapture: integrating subject and style concepts to generate high-resolution portrait images using just a few subject and style references

😎Review https://t.ly/c9rOo
😎Paper https://arxiv.org/pdf/2309.06895.pdf

⚽ Dynamic NeRFs for Soccer ⚽

👉SoccerNeRF: first attempt of "cheap" NeRF applied to football for reconstructing soccer replays in space and time.

😎Review https://t.ly/Ywcvk
😎Paper arxiv.org/pdf/2309.06802.pdf
😎Project https://soccernerfs.isach.be/
😎Code github.com/iSach/SoccerNeRFs

☢️ GlueStick: Graph Neural Matching ☢️

👉GlueStick is joint deep matcher for points and lines that leverages the connectivity information between nodes to better glue them together

😎Review https://t.ly/Atxqo
😎Paper arxiv.org/pdf/2304.02008.pdf
😎Code https://github.com/cvg/GlueStick

🫀CPR-Coach: Neural Cardiopulmonary Resuscitation🫀

👉CPR-Coach: fine-grained action recognition in cardiopulmonary resuscitation

😎Review https://t.ly/Qbg4K
😎Paper arxiv.org/pdf/2309.11718.pdf
😎Code github.com/Shunli-Wang/CPR-Coach
😎Project shunli-wang.github.io/CPR-Coach