project page: https://
vgg-t.github.io
paper：
https://
arxiv.org/abs/2503.11651
code：
https://
github.com/facebookresear
ch/vggt
…
demo：
https://
huggingface.co/spaces/faceboo
k/vggt
…
MACHINE LEARNING
-
VGG-T: Advanced Visual Foundation Model from Facebook Research
By
–
-
Feed-Forward Networks Infer 3D Scene Attributes Directly
By
–
Can a feed-forward neural network directly infer all key 3D attributes of a scene? This CVPR 2025 paper says YES.
— 机器之心 JIQIZHIXIN (@jiqizhixin) 28 mars 2025
The proposed VGGT can directly infer camera parameters, point maps, depth maps, and 3D point tracks from one, a few, or even hundreds of scene views. 🚀🔍 pic.twitter.com/Gm3z2ZAn6RCan a feed-forward neural network directly infer all key 3D attributes of a scene? This CVPR 2025 paper says YES.
The proposed VGGT can directly infer camera parameters, point maps, depth maps, and 3D point tracks from one, a few, or even hundreds of scene views. -
Differences Between AI, Machine Learning, and Deep Learning
By
–
What's the difference between AI, Machine Learning, and Deep Learning?
— God of Prompt (@godofprompt) 27 mars 2025
100% made with AI pic.twitter.com/EvG6QURhmpWhat's the difference between AI, Machine Learning, and Deep Learning? 100% made with AI
-
LLM Capabilities Beyond Next Token Prediction
By
–
There's a lot more happening inside an LLM than just next token prediction
-
AI Model Deployment Challenges for Real-Time Robotics
By
–
One of the very interesting challenges with AI models for robotics is the constraints you're working with when deploying models that need to run in realtime on-device.
-
Keras Neural Network Library Celebrates Ten Year Anniversary
By
–
Happy birthday to Keras, the open-source neural-network library, which was launched 10 years ago today. It's now one of the world’s ten most-used software tools: https://
github.com/keras-team/ker
as
… -
RAG Fine-tuning Hybrid Reduces AI Tutor Costs by 88%
By
–
Excited to share insights from our recent (first ever) guest post on Decoding ML by @iusztinpaul ! We wrote about how a proper RAG + fine-tuning (hybrid) setup can reduce expenses by 88% on our own AI tutor :
-
Amazon Launches AI-Powered Shopping Feature Called Interests
By
–
Amazon launched a new AI-powered shopping feature called "Interests" It will automatically scan the store to notify users about new products based on their shopping prompts, tailored to interests & price preferences Available to select U.S. customers!
-
Claude Sonnet 3.7 MAX on Cursor 500k? Enterprise Only?
By
–
It is yet unclear if this will be offered to Enterprise customers only, as some reports suggest it is already the case. Is Claude Sonnet 3.7 MAX on Cursor 500k? Full scoop
-
Pointwise vs Cumulative Metrics in Language Model Evaluation
By
–
Good catch! The numbers in the old table were pointwise estimates – pointwise performance is a bucketed estimate over context lengths, while the paper reports a cumulative average over context lengths. Pointwise and cumulative metrics are naturally incomparable and the pointwise