Show HN: From Photos to Positions: Prototyping VLM-Based Indoor Maps

#1

Post by accurrent » Sat, Jul 05, 2025, 11:19 PM UTC

Just a fun hack I did while bored over the weekend. My wife was busy shopping, it got me thinking that can VLMs solve the indoor location problem in a mall? Can I just show a VLM a map and an image and have it doa good enough job locating me? I hacked this P.O.C and it seems to work.

Show HN: From Photos to Positions: Prototyping VLM-Based Indoor Maps
arjo129.github.io

Re: Show HN: From Photos to Positions: Prototyping VLM-Based Indoor Maps

#2

Post by rohanrao123 » Tue, Jul 08, 2025, 2:45 AM UTC

Pretty cool! It reminded me of this work from NVIDIA Research - https://nvidia-ai-iot.github.io/remembr where they used VLMs and RAG on top of a real robot to navigate the Voyager campus in Santa Clara. You also might like the new OpenAI o3 models and how well they can play GeoGuessr ;)

https://simonwillison.net/2025/Apr/26/o3-photo-locations, https://news.ycombinator.com/item?id=43835044, https://www.astralcodexten.com/p/testing-ais-geoguessr-geniu...

Re: Show HN: From Photos to Positions: Prototyping VLM-Based Indoor Maps

#3

Post by accurrent » Wed, Jul 09, 2025, 1:51 PM UTC

rohanrao123 wrote: post #2

Pretty cool! It reminded me of this work from NVIDIA Research - https://nvidia-ai-iot.github.io/remembr where they used VLMs and RAG on top of a real robot to navigate the Voyager campus in Santa Clara. You also might like the new OpenAI o3 models and how well they can play GeoGuessr ;) https://simonwillison.net/2025/Apr/26/o3-photo-locations , https://news.ycombinator.com/item?id=43835044 , https://www.astralcodexte…

Yep I've seen the NVidia research stuff. It's pretty cool.