Because of forced 2FA of Github.com this repository moved to https://codeberg.org/SaschaHornauer/Batvision

Please find any future updates there. Old content is kept here to avoid breaking old links.

BatVision: Learning to See 3D Spatial Layout with Two Ears

Code of the original Batvision paper for reference. It is no longer actively maintained but we will add code from the newest paper soon and restructure this repo to see straighforward which code goes with which paper.

We just released the dataset! Please find details how to download in the dataset paper and check out our other related work.

The Audio-Visual BatVision Dataset for Research on Sight and Sound (Arxiv Version)

The Audio-Visual BatVision Dataset for Research on Sight and Sound (IROS published version)

Project Page of the Audio-Visual Batvision Dataset

BatVision: Learning to See 3D Spatial Layout with Two Ears

BatVision with GCC-PHAT Features for Better Sound to Vision Predictions

Note: A bugfix was added in an effort to make the code compatible to be used with Data Distributed Parallel however there is no long term experience with using DDP with this code so please report issues if you find bugs.

Note2: Because of the forced 2FA of Github.com this repository may soon become abandoned. Please see https://amandinebtto.github.io/Batvision-Dataset/ for all new developments.