Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

A Real-Time Wideband Neural Vocoder at 1.6 Kb/S Using LPCNet

people.xiph.org

Post reply on HN
Print view View original

31–40 of 76 posts

Re: A Real-Time Wideband Neural Vocoder at 1.6 Kb/S Using LPCNet

#31

Post by petercooper »

bonzini wrote: post #30
petercooper wrote: post #15

Earlier quoted context omitted.

Here we go! This is the first minute or so of Penny Lane by The Beatles converted down to a 10KB .bin and then back to a .wav: http://no.gd/pennylane.wav .. unsurprisingly the vocals remain recognizable, but the music barely at all.

Can you try it with Tom's Diner by Suzanne Vega? It's sung without any instruments, and an early version of MP3 reportedly was a disaster on that song.

Here you go: http://no.gd/vega2.wav

It holds up ridiculously well considering the entire song compresses down to 25392 bytes.

Re: A Real-Time Wideband Neural Vocoder at 1.6 Kb/S Using LPCNet

#33

Post by m0zg »

rasz wrote: post #5

3 Gflops, we are deep beyond diminishing returns here. Opus seems good enough.

Not really, no. Especially not if this is implemented in a specialized accelerator. A GFLOP is not that much there. Also, like most other neural network algorithms, this could also be done in fixed point, thereby further reducing the computational cost.

Re: A Real-Time Wideband Neural Vocoder at 1.6 Kb/S Using LPCNet

#34

Post by tachyonbeam »

rasz wrote: post #5

3 Gflops, we are deep beyond diminishing returns here. Opus seems good enough.

There are technologies to compress deep networks by pruning weak connections. I don't believe the author is using this, so it's likely the computational cost could be reduced by a factor of 10. It could also be that simple tweaks to the NN architecture also work (was the author aiming for using a network as small as possible to begin with?).

Re: A Real-Time Wideband Neural Vocoder at 1.6 Kb/S Using LPCNet

#35

Post by opportune »

gok wrote: post #8
rasz wrote: post #5

3 Gflops, we are deep beyond diminishing returns here. Opus seems good enough.

3 GFLOP/sec sounds like a lot but it's considerably less math than the radio DSPs inside any modern phone's baseband is doing during a phone call.

I don't know much about phone tech, are the basebands really doing math or just instrumenting? My assumption would be that there is just some sensor writing to a buffer at a high frequency but that whatever processes that buffer operates at a lower frequency.

Re: A Real-Time Wideband Neural Vocoder at 1.6 Kb/S Using LPCNet

#36

Post by jmvalin »

tachyonbeam wrote: post #34
rasz wrote: post #5

3 Gflops, we are deep beyond diminishing returns here. Opus seems good enough.

There are technologies to compress deep networks by pruning weak connections. I don't believe the author is using this, so it's likely the computational cost could be reduced by a factor of 10. It could also be that simple tweaks to the NN architecture also work (was the author aiming for using a network as small as possible to begin with?).

Actually, what's in the demo already includes pruning (through sparse matrices) and indeed, it does keep just 1/10 of the weights as non-zero. In practice it's not quite a 10x speedup because the network has to be a bit bigger to get the same performance. It's still a pretty significant improvement. Of course, the weights are pruned by 16x1 blocks to avoid hurting vectorization (see the first LPCNet paper and the WaveRNN paper for details).

Re: A Real-Time Wideband Neural Vocoder at 1.6 Kb/S Using LPCNet

#37

Post by StavrosK »

petercooper wrote: post #31
bonzini wrote: post #30

Earlier quoted context omitted.

Can you try it with Tom's Diner by Suzanne Vega? It's sung without any instruments, and an early version of MP3 reportedly was a disaster on that song.

Here you go: http://no.gd/vega2.wav It holds up ridiculously well considering the entire song compresses down to 25392 bytes.

The lyrics of the song are 1200 characters long, so this version of the song only takes up twenty times more space than the written lyrics.

Re: A Real-Time Wideband Neural Vocoder at 1.6 Kb/S Using LPCNet

#39

Post by godelski »

petercooper wrote: post #31
bonzini wrote: post #30

Earlier quoted context omitted.

Can you try it with Tom's Diner by Suzanne Vega? It's sung without any instruments, and an early version of MP3 reportedly was a disaster on that song.

Here you go: http://no.gd/vega2.wav It holds up ridiculously well considering the entire song compresses down to 25392 bytes.

I'm getting a 404 on this
Post reply on HN