This media is not supported in your browser
VIEW IN TELEGRAM
Long-term Human Motion Prediction with Scene Context
Forwarded from Neurohive - Нейронные сети, AI, ML, DL
Исследователи из Salesforce обучили языковую модель восстанавливать структуру белка. В качестве архитектуры использовали языковую модель BERT. Помимо модели, разработчики публикую инструмент для трехмерной визуализации механизма внимания из обученной модели.
Forwarded from Artificial Intelligence && Deep Learning (SHOHRUH)
Real time face recognition with Android + TensorFlow Lite
The impressive effect of having the state-of-the-art running on your hands
1. @DeepLearning_ai
2. https://medium.com/@estebanuri/real-time-face-recognition-with-android-tensorflow-lite-14e9c6cc53a5
The impressive effect of having the state-of-the-art running on your hands
1. @DeepLearning_ai
2. https://medium.com/@estebanuri/real-time-face-recognition-with-android-tensorflow-lite-14e9c6cc53a5
Medium
Real time face recognition with Android + TensorFlow Lite
The impressive effect of having the state-of-the-art running on your hands
This media is not supported in your browser
VIEW IN TELEGRAM
A Cordial Sync: Going Beyond Marginal Policies for Multi-Agent Embodied Tasks
This media is not supported in your browser
VIEW IN TELEGRAM
One Policy to Control Them All: Shared Modular Policies for Agent-Agnostic Control
This media is not supported in your browser
VIEW IN TELEGRAM
Style-Controllable Speech-Driven Gesture Synthesis Using Normalising Flows
Media is too big
VIEW IN TELEGRAM
Deep Textures: using machine learning to create unexpected materials with textures and normal data.
This media is not supported in your browser
VIEW IN TELEGRAM
Four-channel synthesis in StyleGAN:
RGB and elevation
RGB and elevation
This media is not supported in your browser
VIEW IN TELEGRAM
Implicit Mesh Reconstruction from Unannotated Image Collections