search

bdaiinstitute / vlfm

The repository provides code associated with the paper VLFM: Vision-Language Frontier Maps for Zero-Shot Semantic Navigation (ICRA 2024)
http://naoki.io/portfolio/vlfm.html
MIT License
194 stars 14 forks source link

Text prompt #31

Closed DoongLi closed 3 months ago

DoongLi commented 3 months ago

Hi, congratulations to this paper for winning ICRA awards. I have a question about Text prompt, is it generated by BLIP-2 or artificially?

DoongLi commented 3 months ago

like：Text prompt: "Seems like there's a microwave ahead."

naokiyokoyama commented 3 months ago

The prompt is a fixed template, meaning it should always just be "Seems like there's a ___ ahead." but the blank is replaced with the target object category. It could be altered to use a different prompt template, but we have found this one to work well.

DoongLi commented 3 months ago

I got it, thx!