MP3 and AAC Explained

By K. Brandenburg. 1999

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Toward Domain-Invariant Speech Recognition via Large Scale Training (2018, methodology): “We train with the MPEG-2 Audio Layer 3 (MP3) and Advanced Audio Coding (AAC) codecs, both of which perform perceptual audio coding [20].”
Distributed Compressive Sensing (2009, unknown): “For example,smoothsignalsaresparseintheFourierbasis,andpiecewisesmoothsignalsaresparse inawaveletbasis[7];thecommercialcodingstandardsMP3[8],JPEG[9],andJPEG2000[10] directlyexploitthissparsity.”
Exploiting time-frequency patterns with LSTM-RNNs for low-bitrate audio restoration (2019, background): “This is commonly due to the fact that the amount of bits required to keep the amplitude of quantization noise below the masked threshold is insufﬁcient, leading to audible noise, so-called spectral holes, or due to signal bandwidth reduction [5, 56].”
Managing performance vs. accuracy trade-offs with loop perforation (2011, unknown): “We note that developers have, for years, manually developed algorithms (most prominently, lossy compression algorithms [36, 7]) that are designed to operate at a variety of points in a performance vs.”