AI Paper Atlas

Research map / Benchmarks / AudioSet

AudioSet leaderboard

10 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on AudioSet. Best mAP: 49.6 (MBT (AV), NeurIPS 2021).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

AudioSet · mAP

10 papers, each with its best reported number for its own method; higher is better.

#mAP ↑MethodPaperTopic
149.6MBT (AV)
ViT-Base
Attention Bottlenecks for Multimodal Fusion
NeurIPS 2021 · AudioSet comparison table
Cross-modal learning
244.9Perceiver IO
Perceiver IO
Perceiver IO: A General Architecture for Structured Inputs & Outputs
ICLR 2022 · AudioSet table
Human and embodied modeling
344.2Perceiver (mel spectrogram, tuned)Perceiver: General Perception with Iterative Attention
ICML 2021 · AudioSet results table
Attention and transformers
439.8SUSTAIN - 2 Teachers
WEANET
A Sequential Self Teaching Approach for Improving Generalization in Sound Event Recognition
ICML 2020 · Table: Comparison with state-o
Self-supervised and contrastive
539.4VATT-BaseVATT: Transformers for Multimodal Self-Supervised Learning from Raw Video, Audio and Text
NeurIPS 2021 · AudioSet and ImageNet results
Action recognition
626.7VIT-LENS-L (I+T)
OpenCLIP ViT-L/14
ViT-Lens: Towards Omni-modal Representations
CVPR 2024 · Table 3
CLIP and zero-shot
726.45InternVL†_IB++ (AT E.)
InternVL†_IB
FreeBind: Free Lunch in Unified Multimodal Space via Knowledge Fusion
ICML 2024 · Results of zero-shot classific
Cross-modal learning
825.57OmniBind-LargeOmniBind: Large-scale Omni Multimodal Representation via Binding Spaces
ICLR 2025 · Zero-shot classification resul
Cross-modal learning
918.7LTU
LLaMA-7B
Listen, Think, and Understand
ICLR 2024 · Closed-ended audio task perfor
AI and society
1018.6MorphNet
AudioResNet
MorphNet: Fast & Simple Resource-Constrained Structure Learning of Deep Networks
CVPR 2018 · Table: Improved Performance at
RNNs and normalization

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.