Dive into the #AIRevolution with #CVPR2023! Experience the fast-paced evolution of AI first-hand and engage with the greatest minds innovating in AI and Computer Vision. At CVPR, you're not just keeping up with the latest tech; you are helping to create it. Let's pioneer the
INNOVATION
-

Drones Reconstruct Ancient Cathedral with Innovative Technology
By
–
Drones complete a cathedral destroyed centuries ago Credit: Cyberdrone Drone Shows
#innovation #techforgood #tech #drones -

CAJun: Continuous Adaptive Jumping Framework for Legged Robots
By
–
CAJun: Continuous Adaptive Jumping using a Learned Centroidal Controller
— AK (@_akhaliq) 19 juin 2023
paper page: https://t.co/9nMSBerv5w
present CAJun, a novel hierarchical learning and control framework that enables legged robots to jump continuously with adaptive jumping distances. CAJun consists of a… pic.twitter.com/5XMbsEFGhSCAJun: Continuous Adaptive Jumping using a Learned Centroidal Controller paper page: https://
huggingface.co/papers/2306.09
557
… present CAJun, a novel hierarchical learning and control framework that enables legged robots to jump continuously with adaptive jumping distances. CAJun consists of a -

Robot Learning with Sensorimotor Pre-training Using Transformer
By
–
Robot Learning with Sensorimotor Pre-training
— AK (@_akhaliq) 19 juin 2023
paper page: https://t.co/JNC4XuEu2f
present a self-supervised sensorimotor pre-training approach for robotics. Our model, called RPT, is a Transformer that operates on sequences of sensorimotor tokens. Given a sequence of camera… pic.twitter.com/DdW8qWTcDPRobot Learning with Sensorimotor Pre-training paper page: https://
huggingface.co/papers/2306.10
007
… present a self-supervised sensorimotor pre-training approach for robotics. Our model, called RPT, is a Transformer that operates on sequences of sensorimotor tokens. Given a sequence of camera -

AvatarBooth: High-Quality Customizable 3D Human Avatar Generation
By
–
AvatarBooth: High-Quality and Customizable 3D Human Avatar Generation
— AK (@_akhaliq) 19 juin 2023
paper page: https://t.co/5WIusjAJDA
introduce AvatarBooth, a novel method for generating high-quality 3D avatars using text prompts or specific images. Unlike previous approaches that can only synthesize… pic.twitter.com/zzfmvxsirbAvatarBooth: High-Quality and Customizable 3D Human Avatar Generation paper page: https://
huggingface.co/papers/2306.09
864
… introduce AvatarBooth, a novel method for generating high-quality 3D avatars using text prompts or specific images. Unlike previous approaches that can only synthesize -

Scaling Open-Vocabulary Object Detection with Vision-Language Models
By
–
Scaling Open-Vocabulary Object Detection paper page: https://
huggingface.co/papers/2306.09
683
… Open-vocabulary object detection has benefited greatly from pretrained vision-language models, but is still limited by the amount of available detection training data. While detection training data can -
MRI reveals complexity of human vocal apparatus and speech mechanics
By
–
MRI scan of a person talking
— Pascal Bornet (@pascal_bornet) 19 juin 2023
Speaking seems so easy, yet our "vocal apparatus" is very complex, involving more than 100 muscles and organs such as lips, tongue, teeth, mouth and throat… Magic of nature!
Credit: Jens Frahm / MPIBPC#biology #technology #innovation pic.twitter.com/9Snle0hIinMRI scan of a person talking Speaking seems so easy, yet our "vocal apparatus" is very complex, involving more than 100 muscles and organs such as lips, tongue, teeth, mouth and throat… Magic of nature! Credit: Jens Frahm / MPIBPC
#biology #technology #innovation -

Block-State Transformer: Advancing State Space Models for Long Sequences
By
–
Block-State Transformer paper page: https://
huggingface.co/papers/2306.09
539
… State space models (SSMs) have shown impressive results on tasks that require modeling long-range dependencies and efficiently scale to long sequences owing to their subquadratic runtime complexity. Originally designed -

Inverse Scaling: When Bigger Isn’t Better in Language Models
By
–
Inverse Scaling: When Bigger Isn't Better paper page: https://
huggingface.co/papers/2306.09
479
… Work on scaling laws has found that large language models (LMs) show predictable improvements to overall loss with increased scale (model size, training data, and compute). Here, we present evidence
