urbanist-ai Goto Github PK

pathplanning

Common used path planning algorithms with animations.

pearl-drl

A Production-ready Reinforcement Learning AI Agent Library brought by the Applied Reinforcement Learning team at Meta.

picogpt

An unnecessarily tiny implementation of GPT-2 in NumPy.

prettymapp

🖼️ Create beautiful maps from OpenStreetMap data in a streamlit webapp

python-visual-odometry

Python implementation of Visual Odometry algorithms from http://rpg.ifi.uzh.ch/

raspi-scale

for my Raspberry Pi coffee container

robustgp

Robust initialisation of inducing points in sparse variational GP regression models.

route-gym

satclip

PyTorch implementation of SatCLIP

slam_3d_plvs

PLVS is a real-time SLAM system with points, lines, volumetric mapping and 3D unsupervised incremental segmentation.

stable-diffusion-tensorflow

Tensorflow port of the Stable Diffusion Model

streamdiffusion

StreamDiffusion: A Pipeline-Level Solution for Real-Time Interactive Generation

streamlit-geospatial

A multi-page streamlit app for geospatial

street_gaussians

Code for "Street Gaussians for Modeling Dynamic Urban Scenes"

svi_to_pointcloud

Openframeworks implementation to create a point cloud from Google Street View

text-generation-webui

A gradio web UI for running Large Language Models like LLaMA, llama.cpp, GPT-J, Pythia, OPT, and GALACTICA.

torch-fidelity

High-fidelity performance metrics for generative models in PyTorch

trafficbots

TrafficBots: Towards World Models for Autonomous Driving Simulation and Motion Prediction. ICRA 2023.

treepedia_public

Treepedia package for public use

urbangiraffe

Official PyTorch implementation of the UrbanGIRAFFE@ICCV2023

urbani_turing_game

This is a social experiment inspired by the Turing test. We ask participants to define whether a given image is a real or fake street-level image. Images appear only for one second twice to make a judgement and in the end, you will get a score of what you have achieved. The average reaction time for humans is 0.25 seconds to a visual stimulus (4 images per second). The synthesis images are generated by a GAN model created by Mohamed R. Ibrahim (Currently under review). The purpose of this game is to validate the model. Nevertheless, it aims to show quantitive evidence of how mixed reality can alter our ground truth. The results of this experiment might be used for AI perception research.

urbanist-ai.github.io

Conference website

worldmodels-a3c

World Models with A3C on Carracing-v0 in gym

ydlidar-sdk

Driver for receiving YD LiDAR data and more...

yolov6

YOLOv6: a single-stage object detection framework dedicated to industrial applications.