M2D

Reported on 5 benchmarks across 3 tasks · 3 papers · 3 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Time Series2 results

Phonocardiogram ClassificationonCirCor DigiScope
Unweighted average recall· 2024-04-26
0.713
SOTA
Exploring Pre-trained General-purpose Audio Representations for Heart Murmur Detection arXiv:2404.17107
Phonocardiogram ClassificationonCirCor DigiScope
Weighted Accuracy· 2024-04-26
0.832
SOTA
Exploring Pre-trained General-purpose Audio Representations for Heart Murmur Detection arXiv:2404.17107

Audio2 results

Emotion RecognitiononEmomusic
EmoA· 2025-03-28
76.1
best: 77.4 (M2D-CLAP)
M2D2: Exploring General-purpose Audio-Language Representations Beyond CLAP arXiv:2503.22104
Emotion RecognitiononEmomusic
EmoV· 2025-03-28
59.4
best: 61.9 (M2D-CLAP)
M2D2: Exploring General-purpose Audio-Language Representations Beyond CLAP arXiv:2503.22104

Speech1 result

Keyword SpottingonGoogle Speech Commands
Google Speech Commands V2 35· 2022-10-26
98.5
best: 99.1 (WaveFormer)
SOTA
Masked Modeling Duo: Learning Representations by Encouraging Both Networks to Model the Input arXiv:2210.14648