Text-to-Image Generation on GenEval (InBSim, DrSim, VD/p, CLIP, HPS) benchmark leaderboard · SOTA2 Research

Text-to-Image Generation on GenEval (InBSim, DrSim, VD/p, CLIP, HPS)

0.634DrSimCADS

Evaluation Results

Method	Links
CADS2026.05		0.634	0.299	3.348	22.82	0.257
NoiseOpt*2026.05		0.446	—	2.753	—	0.293
Input Noise2026.05		0.445	0.592	2.476	27.32	0.287
CADS2026.05		0.431	0.559	2.55	27.34	0.291
STRIDE2026.05		0.423	0.607	2.428	29.04	0.282
GI*2026.05		0.413	—	2.473	—	0.296
SPELL2026.05		0.369	0.651	2.284	27.88	0.272
DDC2026.05		0.368	0.627	2.343	28.52	0.274
Baseline2026.05		0.363	0.669	2.204	28.34	0.296
STRIDE2026.05		0.337	0.679	2.175	28.86	0.287
STRIDE No-PCA2026.05		0.322	0.696	2.123	29.26	0.279
Input Noise2026.05		0.314	0.701	2.087	28.92	0.291
i.i.d.*2026.05		0.307	—	2.013	—	0.304
STRIDE No-PCA2026.05		0.274	0.73	1.986	29.13	0.29
Baseline2026.05		0.264	0.741	1.948	29.2	0.29