💡🔦 DarkLighter, a neural network to see in the dark 💡🔦

👉Current CNN-based trackers can hardly generalize well to low-light scenes that are commonly... 12 comments on LinkedIn

⏹⏹Facebook AI unveils MViT, a novel (and amazing) Transformer architecture for representation learning from visual data⏹⏹

👉MViT it’s a family of visual... 10 comments on LinkedIn

🚤🛳Using a super-resolution #neuralnetwork to hyper-compress video clips for streaming🚤🛳

👉Internet video delivery has undergone a tremendous explosion...

📷🎞Tencent #AI just opened the source code of GFP-GAN, the most advanced photo restoration system ever 📷🎞

👉Blind face restoration usually relies on facial…

🟩🟦TikTok unveils the most advanced, real-time, HD, human video matting ever. Live demo available!🟩🟦

👉Background replacement has many practical… | 10 comments on LinkedIn

🚔🚖From Wuhan, YOLOP: You Only Look Once for Panoptic Driving Perception in #selfdriving 🚔🚖

👉A panoptic driving perception system is an essential part of… | 16 comments on LinkedIn

🍙🍙SenseTime 商汤科技 unveils CSG-Stump Net, an unsupervised #AI for learning how a component is built🍙🍙

👉Generating an interpretable and compact representation...

♻♻Facebook unveils a novel #AI to create immersive #VR experiences by (non immersive) video manipulation♻♻

👉Image view synthesis has seen great success...

🛰🛰Australian researchers open the biggest adversarial dataset to fool #AI satellites🛰🛰

👉Deep neural networks (DNNs) have become essential for processing...

Login to LinkedIn to keep in touch with people you know, share ideas, and build your career.

🍏🍎Apple opens GSN, a generative model which learns to synthesize radiance fields of indoor scenes🍏🍎

👉Spatial understanding entails the ability to... 12 comments on LinkedIn

🌳🌲How to acquire high-res point clouds of the #forest with the Apple #LiDAR in #iPad or #iPhone12 Pro🌳🌲

👉The development of new measurement techniques...

🍭🍭Stanford University unveils π-GAN: a novel Periodic Implicit GAN for 3D-aware image synthesis🍭🍭

👉Existing 3D-aware image synthesis methods fall...

🏌‍♀🏌 SenseTime 商汤科技 unveils a novel multi-level attentions framework for #3D human estimation🏌‍♀🏌

👉3D human shape and pose estimation is the essential… | 20 comments on LinkedIn

🚚🚛 German researchers unveil NEAT, NEural ATtention fields for End-to-End Autonomous Driving 🚚🚛

👉Efficient reasoning about the semantic, spatial, and…

😵🎁 LAION-400M: the world’s largest openly available image-text-pair dataset ever 😵🎁

Highlights:
✅A novel image-text-pair dataset with 400M…

🎯🎯Korean researchers introduce LineTR, a context-aware descriptor for visual localization inspired by #NLP🎯🎯

👉Visual features are widely utilized in…

🌹🌷 Israeli researchers unveil DeepSIM, a neural generative model via single image 🌹🌷

👉#Deeplearning significantly boosted performance on image… | 29 comments on LinkedIn

🍭🍭Researchers from Colombia introduce Panoptic Narrative Grounding to densely segment text and visual objects🍭🍭

👉Vision and language play a key role in…

☝️🟢Air-Drawing with #DeepLearning for collaborative #elearning ☝️🟢

👉This tool uses Deep Learning to help you draw and write with your hand and webcam. A…