List your company
ResearchFormerly WizWand

Image Classification on KITTI

86.4AccuracyMultiple models
7.25627.80348.3568.897Aug 10, 2022Mar 10, 2023Oct 9, 2023May 8, 2024Dec 7, 2024Jul 7, 2025Feb 5, 2026
Updated 8mo ago

Evaluation results

MethodLinks
Multiple modelsBackbone=ViT-L/142022.0886.4
Joint patchingBackbone=ViT-L/142022.0885.1
Multiple modelsBackbone=ViT-B/162022.0884.8
Joint patchingBackbone=ViT-B/162022.0880.5
Multiple modelsBackbone=ViT-B/322022.0877.9
Joint patchingBackbone=ViT-B/322022.0876.2
Seq. patchingBackbone=ViT-L/14, Seed=22022.0875.4
Seq. patchingBackbone=ViT-L/14, Seed=02022.0871.2
CLIPLinear probing=true, Backbone=ViT-B/16, Pre-train=IT35M2022.1069.7
xCLIPLinear probing=true, Backbone=ViT-B/16, Pre-train=IT35M2022.1069.3
Seq. patchingBackbone=ViT-L/14, Seed=12022.0867.4
Seq. patchingBackbone=ViT-B/16, Seed=22022.0864.4
Seq. patchingBackbone=ViT-B/16, Seed=02022.0864.1
nCLIPLinear probing=true, Backbone=ViT-B/16, Pre-train=IT35M2022.1064
Seq. patchingBackbone=ViT-B/32, Seed=02022.0863.3
Seq. patchingBackbone=ViT-B/16, Seed=12022.0859.6
Seq. patchingBackbone=ViT-B/32, Seed=12022.0858.1
Seq. patchingBackbone=ViT-B/32, Seed=22022.0857
Parallel patchingBackbone=ViT-L/142022.0849.6
Parallel patchingBackbone=ViT-B/162022.0848
CLIPPre-training Data=WIT-400M, Backbone=ViT-L/14, Evaluation Protocol=Linear Probe2024.0742.9
Parallel patchingBackbone=ViT-B/322022.0842.1
CLIP+Pre-training Data=WIT-400M, Backbone=ViT-L/14, Evaluation Protocol=Linear Probe2024.0741.9
xCLIPBackbone=ViT-B/16, Pre-trained on=IT35M, Mode=Zero-shot2022.1039.8
nCLIPBackbone=ViT-B/16, Pre-trained on=IT35M, Mode=Zero-shot2022.1035.4
FLIPPre-training Data=LAION-400M, Backbone=ViT-L/14, Evaluation Protocol=Zero-shot classification2024.0734.8
MLCDPre-training Data=LAION-400M, Backbone=ViT-L/14, Evaluation Protocol=Linear Probe2024.0734.5
UNICOMPre-training Data=LAION-400M, Backbone=ViT-L/14, Evaluation Protocol=Linear Probe2024.0733.4
OpenCLIPPre-training Data=LAION-400M, Backbone=ViT-L/14, Evaluation Protocol=Linear Probe2024.0732.5
CLIP-MaptinyImage Encoder=ViT-0.8M/16, Zero-shot=true2026.0232.5
CLIPBackbone=ViT-B/16, Pre-trained on=IT35M, Mode=Zero-shot2022.1031.6
CLIPImage Encoder=ViT-B/16, Zero-shot=true2026.0231.5
UnpatchedBackbone=ViT-B/162022.0830.5
OpenCLIPPre-training Data=LAION-400M, Backbone=ViT-L/14, Evaluation Protocol=Zero-shot classification2024.0729.3
MLCDPre-training Data=LAION-400M, Backbone=ViT-L/14, Evaluation Protocol=Zero-shot classification2024.0728.7
†TinyCLIPImage Encoder=ViT-0.8M/16, Zero-shot=true2026.0228
CLIPPre-training Data=WIT-400M, Backbone=ViT-L/14, Evaluation Protocol=Zero-shot classification2024.0726.9
UnpatchedBackbone=ViT-L/142022.0826
CLIP (reported)Pre-training Data=WIT-400M, Backbone=ViT-L/14, Evaluation Protocol=Zero-shot classification2024.0723.1
UnpatchedBackbone=ViT-B/322022.0822.6
OpenCLIPImage Encoder=ViT-B/16, Zero-shot=true2026.0217
†TinyCLIPImage Encoder=ViT-8M/16, Zero-shot=true2026.0215.3
CLIP-MapsmallImage Encoder=ViT-8M/16, Zero-shot=true2026.0213.2
TinyCLIPImage Encoder=ViT-39M/16, Zero-shot=true2026.0211.1
CLIP-MapbaseImage Encoder=ViT-39M/16, Zero-shot=true2026.0210.3