Giter Site home page Giter Site logo

giacomopiccinini / unsupervisedimagesegmentation Goto Github PK

View Code? Open in Web Editor NEW
2.0 2.0 0.0 45 KB

Unsupervised Image Segmentation using PyTorch

Python 100.00%
deep-learning image-segmentation pytorch segmentation unsupervised unsupervised-segmentation

unsupervisedimagesegmentation's Introduction

Hi!

I am Giacomo, a Machine Learning engineer specialising in Vision and Audio. Currently, I am the CTO at Gemmo AI, a Dublin-based start-up that provides bespoke Deep Learning solutions.

In my previous life I was a String Theorist at Swansea University (Wales), where I learned how to survive daily rains and to efficiently drink beers.

In my spare time, I am trying to develop tools for leveraging Deep Learning to tackle what I believe are interesting problems in the artistic aspects of Vision and Audio.

If you are interested, get in touch with me on X @G_Piccinini_!

Projects

Here is a list of the projects I am currently working on in my spare time. I have ranked them based on difficulty/size, and you can find the corresponding repos here on GitHub. If you like, do feel free to contribute!

Milli-Projects

rush: A rewriting in Rust of some of bash's commands. Mostly a way to learn Rust!

DrumsGPT: Inspired by Andrej Karpathy's NanoGPT lecture, a GPT-like model for writing (read: completing) drum tracks in MIDI format. Currently private repo, as still in the making.

Micro-Projects

giacomo-website: My first (and only) attempt at creating a personal website. Take a look at it here and feel free to scrape the code! Built with NextJS 14 and deployed on Vercel.

Annotate: despite the questionable name, the aim is to build a set of simple yet effective tools to easily annotate immages locally, without resorting to online services which require your data to be made public (or pay yet another subscription). At the moment only supports YOLO-like bounding box annotations. Docs are in the making.

Nano-Projects

audio-graph: API deployed on Modal to create and compare the frequency spectra of two audio files. Ideally useful for ensuring that, when mixing a song, your track is on par with industry standard.

snippet-api: API deployed on Modal to create images of code snippets automatically. Similar (but not as fancy as) Carbon.

imagesum/audiosum/videosum: Rust scripts for analysing metadata of images/audios/videos and provide a summary. Useful when dealing with large datasets. My first attempt at playing around with Rust.

chronopy: pip-installable package to keep track of execution time in Python scripts. Currently downloadable, pip install chronopy. See also this Medium post about it.

Spotilist: Streamlit application accessible here. Use it to find curators of Spotify playlists that include emerging artists. Try and get in touch with them to see your band featured!

unsupervisedimagesegmentation's People

Stargazers

 avatar  avatar

Watchers

 avatar

unsupervisedimagesegmentation's Issues

Segmentation only resulting one label

hi,
thank you for expanding the code for multiple images. I tried using my own datasets containing thin section of rock images where image size is various. However, the training run smoothly and stop at very early epoch due to the labels reach min number of labels.
do you have a suggestion on how to improve the result? the segmentation result contains only single color for every grid.

thank you
Regards,

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.