EURASIP Journal on Audio, Speech, and Music Processing

Table 4 Consensus scores of English-annotated datasets

From: Automated audio captioning: an overview of recent progress and new challenges

	BLEU\(_1\)	BLEU\(_2\)	BLEU\(_3\)	BLEU\(_4\)	ROUGE-L	METEOR	CIDEr	SPICE	SPIDEr	BERTScore	SentenceBERT
AudioCaps	0.65	0.48	0.37	0.29	0.49	0.28	0.90	0.21	0.56	0.52	0.64
Clotho	0.65	0.49	0.38	0.31	0.50	0.30	0.86	0.23	0.54	0.54	0.53
MACS	0.49	0.28	0.16	0.08	0.32	0.18	0.21	0.13	0.17	0.24	0.52

Back to article page