fhoerst/deepxi repository overview

⁠DeepXi: A Deep Learning Approach for SNR Estimation for speech signals

Image for running the DeepXi project code as an docker image. For usage of DeepXi, please visit the official GitHub project page (https://github.com/anicolson/DeepXi⁠). Run the image in the following way to open an interactive terminal: docker container run -it --name deepxi --rm fhoerst/deepxi:latest Then you can copy the files into the respective folders via docker cp command (https://docs.docker.com/engine/reference/commandline/cp/⁠). Afterwards inference can be performed. Please do not forget to copy the files back from the container to your host system after inference is done.

License: MPL-2.0 License

⁠Thanks to

Aaron Nicolson, Kuldip K. Paliwal, Deep learning for minimum mean-square error approaches to speech enhancement, Speech Communication, Volume 111, 2019, Pages 44-55, ISSN 0167-6393, https://doi.org/10.1016/j.specom.2019.06.002⁠.