Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

CLAP: Learning Audio Concepts From Natural Language Supervision

About

Mainstream Audio Analytics models are trained to learn under the paradigm of one class label to many recordings focusing on one task. Learning under such restricted supervision limits the flexibility of models because they require labeled audio for training and can only predict the predefined categories. Instead, we propose to learn audio concepts from natural language supervision. We call our approach Contrastive Language-Audio Pretraining (CLAP), which learns to connect language and audio by using two encoders and a contrastive learning to bring audio and text descriptions into a joint multimodal space. We trained CLAP with 128k audio and text pairs and evaluated it on 16 downstream tasks across 8 domains, such as Sound Event Classification, Music tasks, and Speech-related tasks. Although CLAP was trained with significantly less pairs than similar computer vision models, it establishes SoTA for Zero-Shot performance. Additionally, we evaluated CLAP in a supervised learning setup and achieve SoTA in 5 tasks. Hence, CLAP's Zero-Shot capability removes the need of training with class labels, enables flexible class prediction at inference time, and generalizes to multiple downstream tasks.

Benjamin Elizalde, Soham Deshmukh, Mahmoud Al Ismail, Huaming Wang• 2022

Related benchmarks

TaskDatasetResultRank
Audio ClassificationESC-50
Accuracy96.7
325
Audio ClassificationUrbansound8K
Accuracy84.2
116
Audio ClassificationESC-50 (test)
Accuracy96.7
84
Musical Instrument ClassificationNSynth
Accuracy68.2
75
Audio ClassificationSPC V2
Accuracy96.8
65
Audio ClassificationESC50
Top-1 Acc95.1
64
Keyword SpottingSpeech Commands V2
Accuracy96.8
61
Environmental Sound ClassificationFSD50K
mAP58.6
60
ClassificationAudioSet (test)
mAP5.8
57
Audio ClassificationGTZAN
Accuracy79.3
54
Showing 10 of 100 rows
...

Other info

Follow for update