Video Instance Segmentation On Ovis 1
Metrics
AP50
AP75
AR1
AR10
mask AP
Results
Performance results of various models on this benchmark
Comparison Table
| Model Name | AP50 | AP75 | AR1 | AR10 | mask AP |
|---|---|---|---|---|---|
| dvis-improved-decoupled-framework-for | 62.8 | 37.3 | 15.8 | 42.9 | 37.2 |
| dvis-daq-improving-video-segmentation-via | 83.8 | 62.9 | - | - | 57.1 |
| dvis-improved-decoupled-framework-for | 68.9 | 40.9 | 16.8 | 47.3 | 41.2 |
| instanceformer-an-online-video-instance | 42.5 | 21.61 | 12.9 | 29.3 | 22.8 |
| universal-instance-perception-as-object | 72.5 | 52.2 | - | - | 49.0 |
| robust-online-video-instance-segmentation | 64.7 | 42.6 | 18.4 | 49.1 | 42.6 |
| instanceformer-an-online-video-instance | 40.7 | 18.1 | 12 | 27.1 | 20.0 |
| crossover-learning-for-fast-online-video | 32.7 | 12.1 | - | - | 14.9 |
| boxvis-video-instance-segmentation-with-box | 68.4 | 39.9 | - | - | 40.6 |
| tarvis-a-unified-approach-for-target-based | 52.5 | 30.4 | 15.9 | 39.9 | 31.1 |
| dvis-improved-decoupled-framework-for | 78.9 | 58.5 | - | - | 53.4 |
| temporally-efficient-vision-transformer-for | 34.9 | 15.0 | - | - | 17.4 |
| stc-spatio-temporal-contrastive-learning-for | 33.5 | 13.4 | - | - | 15.5 |
| devis-making-deformable-transformers-work-for | 59.3 | 38.3 | 16.6 | 39.8 | 35.5 |
| minvis-a-minimal-video-instance-segmentation | 61.5 | 41.3 | 18.1 | 43.3 | 39.4 |
| dvis-decoupled-video-instance-segmentation | 75.9 | 53.0 | 19.4 | 55.3 | 49.9 |
| tube-link-a-flexible-cross-tube-baseline-for | 51.5 | 30.2 | 15.5 | 34.5 | 29.5 |
| mask2former-for-video-instance-segmentation | 36.9 | 14.1 | 9.9 | 24.7 | 16.6 |
| dvis-improved-decoupled-framework-for | 72.5 | 55.0 | 20.8 | 54.6 | 49.6 |
| in-defense-of-online-models-for-video | 51.3 | 30 | 15 | 37.5 | 30.2 |
| a-generalized-framework-for-video-instance | 69.2 | 47.8 | 18.9 | 49.0 | 45.4 |
| vita-video-instance-segmentation-via-object | 51.9 | 24.9 | 14.9 | 33.0 | 27.7 |
| novis-a-case-for-end-to-end-near-online-video | 68.3 | 43.8 | 19.4 | 46.9 | 43.5 |
| univs-unified-and-universal-video | - | - | - | - | 41.7 |
| mdqe-mining-discriminative-query-embeddings | 67.8 | 44.3 | 18.3 | 46.5 | 42.6 |
| crossover-learning-for-fast-online-video | 35.5 | 16.9 | - | - | 18.1 |
| spatial-feature-calibration-and-temporal | 35.4 | 15.2 | 8.4 | 23.1 | 17.3 |
| gratt-vis-gated-residual-attention-for-auto | 60.8 | 36.8 | 16.8 | 40.1 | 36.2 |
| tarvis-a-unified-approach-for-target-based | 67.8 | 44.6 | 18.0 | 50.4 | 43.2 |
| refinevis-video-instance-segmentation-with | 70.4 | 48.4 | 19.1 | 51.2 | 46 |
| dvis-decoupled-video-instance-segmentation | 71.9 | 49.2 | 19.4 | 52.5 | 47.1 |
| occluded-video-instance-segmentation | 29.9 | 12.5 | - | - | 14.3 |
| occluded-video-instance-segmentation | 33.9 | 13.1 | - | - | 15.4 |
| novis-a-case-for-end-to-end-near-online-video | 56.2 | 32.6 | 15.7 | 37.1 | 32.7 |
| universal-instance-perception-as-object | 55.5 | 35.6 | - | - | 34.0 |
| context-aware-video-instance-segmentation | 82.6 | 63.5 | 21.2 | 61.8 | 57.1 |
| tarvis-a-unified-approach-for-target-based | 55.0 | 34.4 | 16.1 | 40.9 | 34.0 |
| ctvis-consistent-training-for-online-video | 60.8 | 34.9 | - | - | 35.5 |
| gratt-vis-gated-residual-attention-for-auto | 69.1 | 47.8 | 19.2 | 49.4 | 45.7 |
| d2conv3d-dynamic-dilated-convolutions-for | 33.8 | 13.7 | - | - | 15.2 |
| general-object-foundation-model-for-images | - | 55.5 | - | - | 50.4 |
| in-defense-of-online-models-for-video | 65.7 | 45.2 | 17.9 | 49.6 | 42.6 |
| devis-making-deformable-transformers-work-for | 47.6 | 20.8 | 12.0 | 28.9 | 23.7 |
| ctvis-consistent-training-for-online-video | 71.5 | 47.5 | - | - | 46.9 |