HyperAI초신경

홈 뉴스 최신 연구 논문 튜토리얼 데이터셋 백과사전 SOTA LLM 모델 GPU 랭킹 컨퍼런스

한국어

HyperAI초신경

Few Shot Audio Classification On Esc 50

평가 지표

Top-1 Accuracy(5-Way-1-Shot)

평가 결과

이 벤치마크에서 각 모델의 성능 결과

모델 이름	Top-1 Accuracy(5-Way-1-Shot)	Paper Title	Repository
SimCLR (FSD50K, RN18)	63.40±0.39	MT-SLVR: Multi-Task Self-Supervised Learning for Transformation In(Variant) Representations
SimpleShot CL2N (AST ImageNet - No fine-tune)	60.41 +- 0.41	MetaAudio: A Few-Shot Audio Classification Benchmark
Prototypical Networks (CRNN)	68.83 +- 0.38	MetaAudio: A Few-Shot Audio Classification Benchmark
Meta-Curvature (CRNN)	76.17 +- 0.41	MetaAudio: A Few-Shot Audio Classification Benchmark
MT-SLVR (SimCLR + MLAP) w/ Parallel Adapters (FSD50K, RN18)	69.53±0.39	MT-SLVR: Multi-Task Self-Supervised Learning for Transformation In(Variant) Representations
SimpleShot CL2N (AST ImageNet & AudioSet- No fine-tune)	64.48 +- 0.41	MetaAudio: A Few-Shot Audio Classification Benchmark
MAML (CRNN)	74.66 ± 0.42	MetaAudio: A Few-Shot Audio Classification Benchmark
SimpleShot CL2N (CRNN)	68.82 +-0.39	MetaAudio: A Few-Shot Audio Classification Benchmark
Meta-Baseline (CRNN)	71.72 +- 0.38	MetaAudio: A Few-Shot Audio Classification Benchmark
Multi-Label Augmentation Prediction (FSD50K, RN18)	37.76±0.34	MT-SLVR: Multi-Task Self-Supervised Learning for Transformation In(Variant) Representations

0 of 10 row(s) selected.