Could you please provide the images before resize to 384?

cvlab-stonybrook / LearningToCountEverything

MIT License

357 stars 72 forks source link

Open donggoing opened 11 months ago

donggoing commented 11 months ago

Thanks a lot!

Yiming-M commented 6 months ago

Hi @Viresh-R and @gshoai ,

I have the same question. Is it possible to provide the original dataset?

Reducing image sizes may lose important info in dense areas and thus lead to underperformance.

Many thanks, Yiming