TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Ask HN: Wagt is the best browser based trainable object detection model?

1 pointsby lukasfischerover 2 years ago
I’m new to object detection - and I find it hard to navigate through all the available solutions. Some frameworks are already a couple of years old land&#x2F;or limited in functionality, or not intended to use in the browser.<p>I’d like to build a art&#x2F;painting detector - think of it as a AI guide for museums. You simply point your phone camera to the art piece and additional info pops up (like audio&#x2F;text etc) - in a browser webapp which is accessible through a QR code at the entrance.<p>The object detection algorithm should not only recognize “painting” but also the actual specific painting. As I understand, I need to train a model with the paintings data. I could not find a solution which works browser only and is actually straight forward to use (something like ml5).<p>I was also looking for image hashing algorithms but hashing the paintings “original” picture and the camera feed (with specific angle&#x2F;lights&#x2F;shadow) did not really work out well.<p>Any recommendations?

no comments

no comments