Papers › EfficientLEAF: A Faster LEarnable Audio Frontend of Questionable Use

EfficientLEAF: A Faster LEarnable Audio Frontend of Questionable Use

12 Jul 2022arXiv:2207.05508archive 2025-07-28

Jan Schlüter, Gerald Gutenbrunner

In audio classification, differentiable auditory filterbanks with few parameters cover the middle ground between hard-coded spectrograms and raw audio. LEAF (arXiv:2101.08596), a Gabor-based filterbank combined with Per-Channel Energy Normalization (PCEN), has shown promising results, but is computationally expensive. With inhomogeneous convolution kernel sizes and strides, and by replacing PCEN with better parallelizable operations, we can reach similar results more efficiently. In experiments on six audio classification tasks, our frontend matches the accuracy of LEAF at 3% of the cost, but both fail to consistently outperform a fixed mel filterbank. The quest for learnable audio frontends is not solved.

PaperPDFCode

Code

cpjku/efficientleaf officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Audio ClassificationClassificationInstrument RecognitionPitch ClassificationSpoken language identification

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Audio Classification BirdCLEF 2021 EfficientLEAF (8s) Accuracy 72.2 #1 of 4 Archive leaderboard report
Audio Classification BirdCLEF 2021 EfficientLEAF Accuracy 42.9 #2 of 4 Archive leaderboard report
Audio Classification BirdCLEF 2021 LEAF Accuracy 42.3 #3 of 4 Archive leaderboard report
Audio Classification BirdCLEF 2021 melspect Accuracy 39.9 #4 of 4 Archive leaderboard report
Audio Classification CREMA-D EfficientLEAF Accuracy 60.2 #1 of 3 Archive leaderboard report
Audio Classification CREMA-D melspect Accuracy 58.8 #2 of 3 Archive leaderboard report
Audio Classification CREMA-D LEAF Accuracy 50.2 #3 of 3 Archive leaderboard report
Audio Classification Speech Commands EfficientLEAF Accuracy 95.2 #5 of 7 Archive leaderboard report
Audio Classification Speech Commands LEAF Accuracy 95.1 #6 of 7 Archive leaderboard report
Audio Classification Speech Commands melspect Accuracy 95.1 #7 of 7 Archive leaderboard report
Instrument Recognition NSynth melspect Accuracy 72.1 #5 of 7 Archive leaderboard report
Instrument Recognition NSynth EfficientLEAF Accuracy 71.7 #6 of 7 Archive leaderboard report
Instrument Recognition NSynth LEAF Accuracy 69.2 #7 of 7 Archive leaderboard report
Spoken language identification VoxForge LEAF Accuracy 91.5 #1 of 3 Archive leaderboard report
Spoken language identification VoxForge EfficientLEAF Accuracy 86.6 #2 of 3 Archive leaderboard report
Spoken language identification VoxForge melspect Accuracy 85.6 #3 of 3 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Average PoolingConvolutionDropoutEfficientNet

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections