HyperAI초신경

Image Sentence Alignment On Valse Counting

평가 지표

Accuracy (%)

pairwise accuracy

평가 결과

이 벤치마크에서 각 모델의 성능 결과

			Paper Title
ViLBERT 12-in-1	64.9	76.7	VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena
LXMERT	52.0	62.2	VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena
CLIP	-	62.1	VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena
ViLBERT	50.7	58.6	VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena
GPT2	-	51.6	VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena
GPT1	-	51.2	VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena
VisualBERT	48.3	48.2	VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena

0 of 7 row(s) selected.