Deep Learning Based Voice Extraction and Primary-Ambience Decomposition for Stereo to Surround Upmixing

Paez Amaro RT, Tejeda Ocampo C, Souza Blanes E, Bharitkar S, Madrid Herrera L. Deep Learning Based Voice Extraction and Primary-Ambience Decomposition for Stereo to Surround Upmixing. In: AES Convention 154. Audio Engineering Society; 2023. Paper 62. Available from: https://aes.org/publications/elibrary-page/?id=22087

@inproceedings{PaezAmaro2023_22087,
  author = {Paez Amaro, Ricardo Thaddeus and Tejeda Ocampo, Carlos and Souza Blanes, Ema and Bharitkar, Sunil and Madrid Herrera, Luis},
  title = {{Deep Learning Based Voice Extraction and Primary-Ambience Decomposition for Stereo to Surround Upmixing}},
  booktitle = {AES Convention 154},
  note = {Paper 62},
  year = {2023},
  month = may,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=22087}
}

TY  - CPAPER
TI  - Deep Learning Based Voice Extraction and Primary-Ambience Decomposition for Stereo to Surround Upmixing
AU  - Paez Amaro, Ricardo Thaddeus
AU  - Tejeda Ocampo, Carlos
AU  - Souza Blanes, Ema
AU  - Bharitkar, Sunil
AU  - Madrid Herrera, Luis
T2  - AES Convention 154
M1  - Paper 62
PY  - 2023
DA  - 2023/05/06
UR  - https://aes.org/publications/elibrary-page/?id=22087
PB  - Audio Engineering Society
LA  - en
AB  - Surround systems have gained popularity in home entertainment despite the fact that most of the cinematic content is delivered in two-channel stereo format. Although there are several upmixing options, it has proven challenging to deliver an upmixed signal that approximates the original directionality and timbre intended by the mixing artist. The aim of this work is to design a two-to-five channels upmixer using a novel upmixing strategy combining voice extraction and primary-ambience decomposition. Results from a modified-MUSHRA test show that our proposed upmixer outperforms established alternatives for cinematic upmixing in perceived spatial and timbral quality.
ER  -