Name: Henry Lydecker
Type: User
Company: The University of Sydney
Bio: Data scientist, CV, NLP, ML, Python, R, data vis. Love all things science, with a particular fondness for ecology and public health.
Twitter: hlydecker
Location: Sydney
Henry Lydecker's Projects
Tools for aerial and satellite imagery acquisition, processing, annotation, and analysis using artificial intelligence
Demo workflow for fine tuning an object detection model on a custom dataset and deploying a prediction web app.
A sample app for the Retrieval-Augmented Generation pattern running in Azure, using Azure Cognitive Search for retrieval and Azure OpenAI large language models to power ChatGPT-style and Q&A experiences.
Docker image for badlands
Bitly v4 API
Tools for training and running detectors and classifiers for wildlife images collected from motion-triggered cameras.
CC Search is a search tool for CC-licensed and public domain content across the internet.
Helper for dealing with MS-COCO annotations
Github suggested I should name this "supreme-spork"; while that name is fantastic this repo doesn't deserve it. This is a collection of resources for computer vision annotation.
Useful papers, notebooks, and guides for computer vision work.
Various random computer vision utility scripts, mostly Python or shell.
Powerful and efficient Computer Vision Annotation Tool (CVAT)
PyTorch package for the discrete VAE used for DALL·E.
A Multiclass Weed Species Image Dataset for Deep Learning
DoEgen: A Python Library for Optimised Design of Experiment Generation and Evaluation
Satellite Image Preprocessing for Enhanced Edge Detection using Canny Edge detection against HED NN
Drag & drop UI to build your customized LLM flow
Currently non existent code for the currently theoretical paper "GANdalf: One Shot Face Tolkienization"
Gemfield homework or libgemfield.so
Aesthetically pleasing machine learning metrics powered by ggplot.
Convert geospatial datasets of shapefiles and rasters into COCO format datasets of images and JSON.
Repo for testing ways to selectively remove files from git history
A robot powered training repository :robot:
GPT-3: Language Models are Few-Shot Learners
Scripts to make heaps of random files.
Python implementation of a Hierarchical-Mask R-CNN
Increment-averaged kriging for 3D prediction of soil properties
Pytorch implementation for "Iterative Human and Automated Identification of Wildlife Images" (Nature -Machine Intelligence, 2021)
Convert JSON annotations into YOLO format.