Startseite Neuigkeiten Neueste Forschungsarbeiten Tutorials Datensätze Wiki SOTA LLM-Modelle GPU-Rangliste Veranstaltungen

Deutsch

Audio Captioning On Audiocaps

Metriken

CIDEr

METEOR

SPICE

SPIDEr

Ergebnisse

Leistungsergebnisse verschiedener Modelle zu diesem Benchmark

Modellname	CIDEr	METEOR	SPICE	SPIDEr	Paper Title	Repository
EnCLAP-large	0.8029	0.2554	0.1879	0.4954	EnCLAP: Combining Neural Audio Codec and Audio-Text Joint Embedding for Automated Audio Captioning
EnCLAP++-large	0.823	0.269	0.197	0.510	EnCLAP++: Analyzing the EnCLAP Framework for Optimizing Automated Audio Captioning Performance
VAST	0.781	0.247	-	-	VAST: A Vision-Audio-Subtitle-Text Omni-Modality Foundation Model and Dataset
AL-MixGen	0.755	-	0.177	0.466	Exploring Train and Test-Time Augmentations for Audio-Language Learning	-
SLAM-AAC	0.841	0.268	0.194	0.518	SLAM-AAC: Enhancing Audio Captioning with Paraphrasing Augmentation and CLAP-Refine through LLMs
LOAE	0.816	0.267	0.193	0.505	Enhancing Automated Audio Captioning via Large Language Models with Optimized Audio Encoding
TopDown-AlignedAtt (1NN)	0.593	-	0.144	0.369	AudioCaps: Generating Captions for Audios in The Wild	-
AL-MixGen + Multi-TTA	0.769	-	0.181	0.475	-	-
EnCLAP++-base	0.815	0.257	0.188	0.501	EnCLAP++: Analyzing the EnCLAP Framework for Optimizing Automated Audio Captioning Performance
CNext-trans	0.8061	0.2527	0.1841	0.4951	-	-
BART + YAMNet + PANNs	0.753	-	0.176	0.465	AUTOMATED AUDIO CAPTIONING BY FINE-TUNING BART WITH AUDIOSET TAGS
EnCLAP-base	0.7795	0.2473	0.1863	0.4829	EnCLAP: Combining Neural Audio Codec and Audio-Text Joint Embedding for Automated Audio Captioning
CNN+Transformer	0.693	-	0.159	0.426	Audio Captioning Transformer
VALOR	0.741	0.231	-	-	VALOR: Vision-Audio-Language Omni-Perception Pretraining Model and Dataset
AutoCap	0.832	0.253	0.182	0.507	Taming Data and Transformers for Audio Generation
Rethink-ACT (AST + TF + MIL)	0.764	0.242	0.180	0.472	Rethinking Transfer and Auxiliary Learning for Improving Audio Captioning Transformer	-

0 of 16 row(s) selected.