HyperAI

Semi Supervised Video Object Segmentation On 21

Metrics

F
FPS
J
Ju0026F

Results

Performance results of various models on this benchmark

Model Name
F
FPS
J
Ju0026F
Paper TitleRepository
DEVA (no OVIS)64.325.355.860.0Tracking Anything with Decoupled Video Segmentation
Cutie (small, MEGA)72.945.564.368.6Putting the Object Back into Video Object Segmentation
DeAOT63.8-55.159.4Decoupling Features in Hierarchical Propagation for Video Object Segmentation
RDE52.9-44.648.8Recurrent Dynamic Embedding for Video Object Segmentation
XMem62.0-53.357.6XMem: Long-Term Video Object Segmentation with an Atkinson-Shiffrin Memory Model
Cutie+ (base, MEGA)75.817.967.671.7Putting the Object Back into Video Object Segmentation
AOT61.3-53.157.2Associating Objects with Transformers for Video Object Segmentation
Cutie (base)67.936.460.064.0Putting the Object Back into Video Object Segmentation
SWEM54.9-46.850.9SWEM: Towards Real-Time Video Object Segmentation with Sequential Weighted Expectation-Maximization
Cutie+ (small, MEGA)74.520.666.070.3Putting the Object Back into Video Object Segmentation
DEVA (with OVIS)70.825.362.366.5Tracking Anything with Decoupled Video Segmentation
Cutie (small)66.245.558.262.2Putting the Object Back into Video Object Segmentation
STCN55.0-46.650.8Rethinking Space-Time Networks with Improved Memory Coverage for Efficient Video Object Segmentation
Cutie (base, with mose)72.336.464.268.3Putting the Object Back into Video Object Segmentation
Cutie (base, MEGA)74.136.465.869.9Putting the Object Back into Video Object Segmentation
Cutie (small, with mose)71.745.563.167.4Putting the Object Back into Video Object Segmentation
SAM2---77.9SAM 2: Segment Anything in Images and Videos
0 of 17 row(s) selected.