Visual Object Tracking On Davis 2016
評価指標
F-measure (Decay)
F-measure (Mean)
F-measure (Recall)
Ju0026F
Jaccard (Decay)
Jaccard (Mean)
Jaccard (Recall)
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
比較表
| モデル名 | F-measure (Decay) | F-measure (Mean) | F-measure (Recall) | Ju0026F | Jaccard (Decay) | Jaccard (Mean) | Jaccard (Recall) |
|---|---|---|---|---|---|---|---|
| cnn-in-mrf-video-object-segmentation-via | 14.7 | 85.0 | 92.1 | 84.2 | 12.3 | 83.4 | 94.9 |
| xmem-long-term-video-object-segmentation-with | - | 91.9 | - | 90.8 | - | 89.6 | - |
| look-before-you-match-instance-understanding | - | 94.2 | - | 93.4 | - | 92.5 | - |
| associating-objects-with-transformers-for | - | 93.3 | - | 92.0 | - | 90.7 | - |
| video-object-segmentation-with-language | 8.6 | 84.2 | 93.9 | 83.65 | 6.9 | 83.1 | 95.7 |
| xmem-long-term-video-object-segmentation-with | - | 93.5 | - | 92.7 | - | 92.0 | - |
| hierarchical-memory-matching-network-for | - | 92.0 | - | 90.8 | - | 89.6 | - |
| associating-objects-with-transformers-for | - | 91.1 | - | 90.4 | - | 89.6 | - |
| collaborative-video-object-segmentation-by | - | 90.5 | - | 89.4 | - | 88.3 | - |
| associating-objects-with-transformers-for | - | 91.1 | - | 89.9 | - | 88.7 | - |
| fast-and-accurate-online-video-object | 5.5 | 79.5 | 89.4 | 80.95 | 4.5 | 82.4 | 96.5 |
| decoupling-features-in-hierarchical | - | 92.5 | - | 91.0 | - | 89.4 | - |
| premvos-proposal-generation-refinement-and | 9.8 | 88.6 | 94.7 | 86.75 | 8.8 | 84.9 | 96.1 |
| siam-r-cnn-visual-tracking-by-re-detection | 4.0 | 80.4 | 87.6 | 78.6 | 2.2 | 76.8 | 86.4 |
| associating-objects-with-scalable | - | 93.6 | - | 92.1 | - | 90.6 | - |
| mobilevos-real-time-video-object-segmentation | - | 91.6 | - | 90.6 | - | 89.7 | - |
| learning-video-object-segmentation-from | 9.0 | 75.4 | 87.1 | 77.55 | 8.9 | 79.7 | 93.1 |
| a-spectral-approach-to-unsupervised-object | - | - | - | - | - | 86.3 | - |
| fast-video-object-segmentation-by-reference | 10.1 | 82.0 | 90.8 | 81.75 | 10.9 | 81.5 | 91.7 |
| associating-objects-with-scalable | - | 94.1 | - | 92.4 | - | 90.6 | - |
| modular-interactive-video-object-segmentation | 5.1 | 92.4 | 96.4 | 91.0 | 6.6 | 89.7 | 97.5 |
| xmem-long-term-video-object-segmentation-with | - | 88.9 | - | 87.8 | - | 86.7 | - |
| online-adaptation-of-convolutional-neural | 5.8 | 84.9 | 89.7 | 85.5 | 5.2 | 86.1 | 96.1 |
| kernelized-memory-network-for-video-object | - | 91.5 | - | 90.5 | - | 89.5 | - |
| feelvos-fast-end-to-end-embedding-learning | 14.1 | 82.2 | 86.6 | 81.65 | 13.7 | 81.1 | 90.5 |
| video-object-segmentation-without-temporal | 8.2 | 87.5 | 95.9 | 86.55 | 5.5 | 85.6 | 96.8 |
| lucid-data-dreaming-for-video-object | 9.7 | 82.0 | 88.1 | 82.95 | 9.1 | 83.9 | 95.0 |
| reliable-propagation-correction-modulation | - | 94 | - | 90.6 | - | 87.1 | - |
| a-generative-appearance-model-for-end-to-end | 9.8 | 82.2 | 90.3 | 81.85 | 9.4 | 81.5 | 93.6 |
| fast-online-object-tracking-and-segmentation | 2.1 | 67.8 | 79.8 | 69.75 | 3.0 | 71.7 | 86.8 |
| separable-structure-modeling-for-semi | 5.6 | 85.6 | 92.3 | 85.9 | 5.3 | 86.2 | 97.1 |
| associating-objects-with-scalable | - | 94.5 | - | 93.0 | - | 91.5 | - |
| decoupling-features-in-hierarchical | - | 93.7 | - | 92.0 | - | 90.3 | - |
| bilateral-space-video-segmentation | 21.3 | 58.8 | 67.9 | 59.4 | 28.9 | 60.0 | 66.9 |
| decoupling-features-in-hierarchical | - | 94.7 | - | 92.9 | - | 91.1 | - |
| xmem-long-term-video-object-segmentation-with | - | 94.4 | - | 93.3 | - | 92.2 | - |
| decoupling-features-in-hierarchical | - | 89.9 | - | 88.9 | - | 87.8 | - |
| ranet-ranking-attention-network-for-fast | 8.2 | 87.6 | 96.1 | 87.1 | 7.4 | 86.6 | 97 |
| pixel-level-matching-for-video-object | 14.7 | 62.5 | 73.2 | 66.35 | 11.2 | 70.2 | 86.3 |
| swem-towards-real-time-video-object-1 | - | 89.0 | - | 88.1 | - | 87.3 | - |
| make-one-shot-video-object-segmentation-1 | - | 87.0 | - | 86.8 | 4.5 | 86.6 | - |
| decoupling-features-in-hierarchical | - | 94.0 | - | 92.3 | - | 90.5 | - |
| xmem-long-term-video-object-segmentation-with | - | 93.2 | - | 92.0 | - | 90.7 | - |
| spatiotemporal-cnn-for-video-object | - | 83.8 | - | 83.8 | - | 83.8 | - |
| region-aware-video-object-segmentation-with | - | 92.6 | - | 91.7 | - | 90.8 | - |
| mobilevos-real-time-video-object-segmentation | - | 92.6 | - | 91.4 | - | 90.3 | - |
| trickvos-a-bag-of-tricks-for-video-object | - | 93.1 | - | 91.8 | - | 90.5 | - |
| 190408141 | 9.0 | 89.5 | 95.5 | 88.55 | 6.9 | 87.6 | 97.3 |
| blazingly-fast-video-object-segmentation-with | 7.8 | 79.3 | 93.4 | 77.4 | 8.5 | 75.5 | 89.6 |
| video-segmentation-via-object-flow | 27.2 | 63.4 | 70.4 | 65.7 | 26.4 | 68.0 | 75.6 |
| online-video-object-segmentation-via | 12.9 | 69.3 | 79.6 | 71.4 | 15.6 | 73.5 | 87.4 |
| learning-fast-and-robust-target-models-for | - | - | - | 81.7 | - | - | - |
| collaborative-video-object-segmentation-by-1 | - | 91.1 | - | 89.9 | - | 88.7 | - |
| associating-objects-with-transformers-for | - | 92.1 | - | 91.1 | - | 90.1 | - |
| xmem-long-term-video-object-segmentation-with | - | 92.7 | - | 91.5 | - | 90.4 | - |
| video-propagation-networks | 14.4 | 65.6 | 69.0 | 67.9 | 12.4 | 70.2 | 82.3 |
| one-shot-video-object-segmentation | 15.0 | 80.6 | 92.6 | 80.2 | 14.9 | 79.8 | 93.6 |
| learning-quality-aware-dynamic-memory-for | - | 93.2 | - | 92.0 | - | 90.7 | - |
| fully-connected-object-proposals-for-video | -1.1 | 49.2 | 49.5 | 53.8 | -2.0 | 58.4 | 71.5 |
| learning-video-object-segmentation-from-2 | 27.2 | 63.6 | 67.7 | 64.65 | 26.4 | 65.7 | 77.7 |
| lsmvos-long-short-term-similarity-matching | 4.9 | 87.3 | 96.1 | 86.5 | 5.1 | 85.7 | 97.1 |
| associating-objects-with-scalable | - | 94.2 | - | 92.4 | - | 90.5 | - |
| video-object-segmentation-using-space-time | 4.2 | 90.1 | 95.2 | 89.4 | 5.0 | 88.7 | 97.4 |
| segflow-joint-learning-for-video-object | 10.4 | 76.0 | 85.5 | 76.05 | 12.1 | 76.1 | 90.6 |
| decoupling-features-in-hierarchical | - | 90.9 | - | 89.3 | - | 87.6 | - |
| associating-objects-with-scalable | - | 93.4 | - | 92.0 | - | 90.5 | - |
| associating-objects-with-transformers-for | - | 90.2 | - | 89.4 | - | 88.6 | - |
| ranet-ranking-attention-network-for-fast | 5.1 | 85.4 | 94.9 | 85.45 | 6.2 | 85.5 | 97.2 |
| trickvos-a-bag-of-tricks-for-video-object | - | 89.9 | - | 89.3 | - | 88.7 | - |
| rethinking-space-time-networks-with-improved | 4.3 | 93.0 | 97.1 | 91.7 | 4.1 | 90.4 | 98.1 |
| crvos-clue-refining-network-for-video-object | 8.8 | 81.0 | 90.3 | 81.6 | 10.0 | 82.2 | 93.9 |
| efficient-video-object-segmentation-via | 10.6 | 72.9 | 84.0 | 73.45 | 9.0 | 74.0 | 87.6 |
| associating-objects-with-scalable | - | 94.4 | - | 93.0 | - | 91.6 | - |
| associating-objects-with-scalable | - | 90.9 | - | 90.3 | - | 89.6 | - |
| an-efficient-3d-cnn-for-actionobject | 4.9 | 77.2 | 84.7 | 77.75 | 2.3 | 78.3 | 91.1 |
| associating-objects-with-transformers-for | - | 87.4 | - | 86.8 | - | 86.1 | - |
| efficient-regional-memory-network-for-video | - | 88.7 | - | 88.8 | - | 88.9 | - |
| fast-video-object-segmentation-with-temporal-1 | - | 68.9 | - | 68.8 | - | 68.6 | - |