High Fidelity Neural Audio Compression · HackerPapers discussions

High Fidelity Neural Audio Compression — mentioned by bravura

Tue, 13 Aug 2024 16:27:28 GMT

For instance: https://github.com/descriptinc/descript-audio-codec/blob/mai... https://github.com/NVIDIA/BigVGAN/blob/main/loss.py#L23 https://arxiv.org/pdf/2210.13438 (the github repo doesn't include training, just inference) It is INCREDIBLY common to use multi-scale spectral loss as the audio distance / objective measure in audio generation. They have some issues (i.e. they aren't always well correlated with human perception) but they are the known-current-best.

Paper and discussion history. A mention is not an endorsement.

High Fidelity Neural Audio Compression — mentioned by tough

Thu, 03 Nov 2022 14:28:05 GMT

> I am curious how they achieve this kind of data compression. Probably ML https://arxiv.org/abs/2210.13438

Paper and discussion history. A mention is not an endorsement.

High Fidelity Neural Audio Compression — HN submission

Thu, 03 Nov 2022 07:57:04 GMT

High Fidelity Neural Audio Compression

3 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

High Fidelity Neural Audio Compression — mentioned by tough

Tue, 01 Nov 2022 17:45:48 GMT

I think they must be using similar AI / ML stuff to this recent work for audio https://arxiv.org/abs/2210.13438 but for video

Paper and discussion history. A mention is not an endorsement.

High Fidelity Neural Audio Compression — mentioned by telotortium

Mon, 31 Oct 2022 04:38:31 GMT

Impressive results - on CD-quality audio too, not just speech as with many recent codecs. One thing I was not able to find, either in the blog post or the paper (https://arxiv.org/pdf/2210.13438.pdf), was the size of the model. EDIT: I downloaded the model from the code (https://github.com/facebookresearch/encodec/blob/3837f0db2a3...). It looks to have 15M parameters in total (this is a CPU-only model, by the way). For inference, I assume we can get away with 16-bit floats, so this clocks in at about 30 MiB. Not particularly favorable compared with LAME on my system (164 KiB), but not at all…

Paper and discussion history. A mention is not an endorsement.

High Fidelity Neural Audio Compression — HN submission

Tue, 25 Oct 2022 18:16:26 GMT

High Fidelity Neural Audio Compression

4 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.