arnavkj1995/VSG
Learning Robust Dynamics Through Variational Sparse Gating
This project helps researchers in artificial intelligence develop more efficient and robust 'world models' for agents in complex environments. It takes in visual sensory data from simulations (like the BringBackShapes environment or DeepMind Control Suite) and outputs improved models that can better predict future outcomes. AI researchers and practitioners working on embodied agents or reinforcement learning would use this to build agents capable of planning in dynamic, multi-object settings.
No commits in the last 6 months.
Use this if you are an AI researcher experimenting with world models and need to improve an agent's ability to learn dynamics in environments with many interacting objects or partial observability.
Not ideal if you are looking for a pre-trained, ready-to-deploy agent for a specific task, or if your primary interest is in environments with very few, simple objects.
Stars
20
Forks
1
Language
Python
License
MIT
Category
Last pushed
Oct 19, 2022
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/arnavkj1995/VSG"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
StanfordASL/Trajectron
Code accompanying "The Trajectron: Probabilistic Multi-Agent Trajectory Modeling with Dynamic...
StanfordASL/Trajectron-plus-plus
Code accompanying the ECCV 2020 paper "Trajectron++: Dynamically-Feasible Trajectory Forecasting...
uber-research/LaneGCN
[ECCV2020 Oral] Learning Lane Graph Representations for Motion Forecasting
agrimgupta92/sgan
Code for "Social GAN: Socially Acceptable Trajectories with Generative Adversarial Networks",...
devendrachaplot/Neural-SLAM
Pytorch code for ICLR-20 Paper "Learning to Explore using Active Neural SLAM"