Image Classification on 13 Diverse Datasets (CIFAR10, ImageNet, DTD, etc.) benchmark leaderboard · SOTA2 Research

Image Classification on 13 Diverse Datasets (CIFAR10, ImageNet, DTD, etc.)

98.5CIFAR10 AccuracyUNICOM

Evaluation Results

Method	Links
UNICOM2023.04		98.5	90.8	95.7	94.6	99.3	93.4	82.4	80	82.2	74.5	94.2	96.7	81.8	89.5
UNICOM2023.04		98.5	90.7	95.7	95.1	99.4	94.3	85.1	79.7	82	78.1	94.5	97.2	82.7	90.2
CLIP-R2023.04		98.1	87.2	96	90.7	99.2	95.3	77.8	81.5	80.9	68	94.9	96.7	84.1	88.5
CLIP2023.04		98	87.5	96.5	90.9	99.2	95.2	77	81.8	82.1	69.4	95.1	98.2	83.9	88.8
CLIP2023.04		97.9	87.4	96	91.5	99.2	95.9	79.9	82.2	83	71.6	95.1	98.1	85.4	89.5
OPEN-CLIP2023.04		97.9	87.9	95.5	93.6	98.8	93.3	78	81	83	64.4	93.3	97.1	81.5	88.1
CLIP-R2023.04		97.8	87.1	96.3	91.4	99.2	95.9	80.9	82.2	82.4	71.2	95.1	96.8	84.9	89.3
UNICOM2023.04		97.3	87.7	95.1	94.3	98.9	91.2	79.3	77.1	81.2	73.4	93.9	97	79.1	88.1
UNICOM2023.04		96.8	86.6	94.6	93.3	98.5	85.8	70.2	74.6	78	70.7	93.1	96.8	75	85.7
OPEN-CLIP2023.04		96.4	84	94.1	91.8	98.1	90.7	71.2	78.7	81.6	59.3	90	96.2	78.5	85.4
CLIP2023.04		96.2	83.1	94.7	86.7	98.1	92.8	67.8	78.4	79.2	59.5	93.1	97.1	80.2	85.1
CLIP-R2023.04		96	82.5	94.1	86	97.8	92.7	61.1	79.1	78.4	58.9	93	95.3	75.3	83.9
CLIP-R2023.04		96	82.5	94.1	86	97.8	92.7	69.5	79.1	78.4	58.9	93	95.3	79.6	84.8
OPEN-CLIP2023.04		95.3	82.2	93.3	87.5	96.5	86.2	61.4	75.3	78.8	52.4	88	96.5	73.8	82.1
CLIP2023.04		95.1	80.5	93	81.8	96.9	88.8	58.5	76.6	76.5	52	90	97	76.1	81.8