MIT-SPARK / Kimera-Semantics

Real-Time 3D Semantic Reconstruction from 2D data
BSD 2-Clause "Simplified" License
631 stars 134 forks source link

Why can't I achieve the effect of the paper after rebuilding the map? #74

Closed HongLouyemeng closed 2 years ago

HongLouyemeng commented 2 years ago

image The above is the result of my reconstruction using "uhumans_12" data, and the following is the result in the paper: Is it the problem of data set or parameters? Can you give me some suggestions? image

ToniRV commented 2 years ago

@HongLouyemeng the figure in the paper is from the uHumans1 dataset: http://mit.edu/sparklab/datasets/uHumans/

https://drive.google.com/file/d/1c6vAHmCd9onycQnCVnw6rK5mNPJ-dC3j/view?usp=sharing

HongLouyemeng commented 2 years ago

@HongLouyemeng the figure in the paper is from the uHumans1 dataset: http://mit.edu/sparklab/datasets/uHumans/

https://drive.google.com/file/d/1c6vAHmCd9onycQnCVnw6rK5mNPJ-dC3j/view?usp=sharing

Thank you for your reply