List your company
ResearchFormerly WizWand

Reasoning Video Object Segmentation on ReVOS 1.0 (test)

0.502Jaccard (J)HyperSeg-3B
0.103680.207090.31050.41391Jul 16, 2024Aug 7, 2024Aug 29, 2024Sep 20, 2024Oct 12, 2024Nov 3, 2024Nov 26, 2024
Updated 8mo ago

Evaluation results

MethodLinks
HyperSeg-3BBackbone=Swin-B, Pre-trained SAM=false2024.110.5020.5580.53
VISABackbone=Chat-UniVi-13B, Instruction Tuning=true2024.070.420.4670.443
VISABackbone=LLaVA-13B, Instruction Tuning=true2024.070.4190.4650.442
VISABackbone=Chat-UniVi-7B, Instruction Tuning=true2024.070.4060.4540.43
VISABackbone=LLaVA-7B, Instruction Tuning=true2024.070.4050.4580.432
VISABackbone=Chat-UniVi-13B, Instruction Tuning=false2024.070.3830.4350.409
VISA-13BBackbone=ViT-H, Pre-trained SAM=true2024.110.3830.4350.409
TrackGPTBackbone=LLaVA-13B, Instruction Tuning=true, Reproduced=true2024.070.3810.4290.405
TrackGPT-13BBackbone=ViT-H, Pre-trained SAM=true2024.110.3810.4290.405
TrackGPTBackbone=LLaVA-7B, Instruction Tuning=true, Reproduced=true2024.070.3680.4120.39
VISABackbone=Chat-UniVi-7B, Instruction Tuning=false2024.070.3670.4170.392
VISA-7BBackbone=ViT-H, Pre-trained SAM=true2024.110.3670.4170.392
LISABackbone=LLaVA-13B, Instruction Tuning=false, Reproduced=true2024.070.3430.3910.367
LISABackbone=LLaVA-7B, Instruction Tuning=false2024.070.3380.3840.361
LISA-7BBackbone=ViT-H, Pre-trained SAM=true2024.110.3380.3840.361
ReferFormerBackbone=Video-Swin-B, Instruction Tuning=false2024.070.2130.2560.234
ReferFormerBackbone=Video-Swin-B, Pre-trained SAM=false2024.110.2130.2560.234
MTTRBackbone=Video-Swin-T, Instruction Tuning=false2024.070.2040.2150.21
LMPMBackbone=Swin-T, Instruction Tuning=false2024.070.1330.2430.188
LMPMBackbone=Swin-T, Pre-trained SAM=false2024.110.1330.2430.188
LLaMA-VID+LMPMBackbone=Swin-T, Instruction Tuning=false2024.070.1280.2370.182
ReferFormerBackbone=ResNet-50, Instruction Tuning=false2024.070.1190.1380.128