Video Summarization on SumMe
64.2F1 Score (Avg)Human
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| HumanEvaluation protocol=Canonical2018.12 | 64.2 | — | |
| CSNetLearning mode=Unsupervised2018.11 | 51.3 | — | |
| CSNetLearning Strategy=Unsupervised2018.11 | 51.3 | — | |
| SUM-FCNAugmented Dataset=true2018.12 | 51.1 | — | |
| VASNetEvaluation protocol=Augmented2018.12 | 51.09 | — | |
| VJMHTSupervision Setting=Supervised2021.12 | 50.6 | — | |
| GLRPESupervision Setting=Supervised2021.12 | 50.2 | — | |
| VASNetEvaluation protocol=Canonical2018.12 | 49.71 | — | |
| CSNetsupLearning Setting=Supervised, Evaluation Protocol=Canonical2018.11 | 48.6 | — | |
| CSNetsupLearning Strategy=Supervised2018.11 | 48.6 | — | |
| CSNetSupervision Setting=Supervised2021.12 | 48.6 | — | |
| SUM-FCNAugmented Dataset=false2018.12 | 47.5 | — | |
| UnpairedVSNSupervision Setting=Unsupervised2021.12 | 47.5 | — | |
| SUM-FCNSupervision Setting=Supervised2021.12 | 47.5 | — | |
| ACGANSupervision Setting=Supervised2021.12 | 47.2 | — | |
| VJMHTSupervision Setting=Unsupervised2021.12 | 47.1 | — | |
| M-AVSEvaluation protocol=Augmented2018.12 | 46.1 | — | |
| ACGANSupervision Setting=Unsupervised2021.12 | 46 | — | |
| SASUMsupEvaluation protocol=Canonical2018.12 | 45.3 | — | |
| RSGNSupervision Setting=Supervised2021.12 | 45 | — | |
| re-SEQ2SEQAugmented Dataset=true2018.12 | 44.9 | — | |
| M-AVSEvaluation protocol=Canonical2018.12 | 44.4 | — | |
| HSA-RNNAugmented Dataset=false2018.12 | 44.1 | — | |
| HMTsplit=canonical2021.09 | 44.1 | — | |
| HMT2021.09 | 44.1 | — | |
| DR-DSNsupEvaluation Protocol=Augmented2017.12 | 43.9 | — | |
| DR-DSNsupEvaluation protocol=Augmented2018.12 | 43.9 | — | |
| GANsupEvaluation Protocol=Augmented2017.12 | 43.6 | — | |
| SUM-GANsupEvaluation protocol=Augmented2018.12 | 43.6 | — | |
| SUM-GANAugmented Dataset=true2018.12 | 43.6 | — | |
| WS-HRLsplit=canonical2021.09 | 43.6 | — | |
| WS-HRLSupervision Setting=Unsupervised2021.12 | 43.6 | — | |
| GANdppEvaluation Protocol=Augmented2017.12 | 43.4 | — | |
| DPP-LSTMEvaluation Protocol=Augmented2017.12 | 42.9 | — | |
| dppLSTMEvaluation protocol=Augmented2018.12 | 42.9 | — | |
| dppLSTMAugmented Dataset=true2018.12 | 42.9 | — | |
| DR-DSNEvaluation Protocol=Augmented2017.12 | 42.8 | — | |
| PCDLSupervision Setting=Unsupervised2021.12 | 42.7 | — | |
| DR-DSNsupEvaluation Protocol=Transfer2017.12 | 42.6 | — | |
| re-SEQ2SEQSupervision Setting=Supervised2021.12 | 42.5 | — | |
| DR-DSNEvaluation Protocol=Transfer2017.12 | 42.4 | — | |
| HSA-RNNsplit=canonical2021.09 | 42.3 | — | |
| RSGNSupervision Setting=Unsupervised2021.12 | 42.3 | — | |
| HSA-RNNSupervision Setting=Supervised2021.12 | 42.3 | — | |
| DR-DSNsupEvaluation Protocol=Canonical2017.12 | 42.1 | — | |
| DR-DSNsupLearning Setting=Supervised2017.12 | 42.1 | — | |
| DR-DSNsupLearning Setting=Supervised, Evaluation Protocol=Canonical2018.11 | 42.1 | — | |
| DR-DSNsupLearning Strategy=Supervised2018.11 | 42.1 | — | |
| DR-DSNsupEvaluation protocol=Canonical2018.12 | 42.1 | — | |
| H-RNNsplit=canonical2021.09 | 42.1 | — | |
| DR-DSNsupsplit=canonical2021.09 | 42.1 | — | |
| H-RNNSupervision Setting=Supervised2021.12 | 42.1 | — | |
| Cycle-SUMLearning Approach=Unsupervised2019.04 | 41.9 | — | |
| DPP-LSTMEvaluation Protocol=Transfer2017.12 | 41.8 | — | |
| GANsupEvaluation Protocol=Canonical2017.12 | 41.7 | — | |
| GANsupLearning Setting=Supervised2017.12 | 41.7 | — | |
| GANsupLearning Setting=Supervised, Evaluation Protocol=Canonical2018.11 | 41.7 | — | |
| GANsupLearning Strategy=Supervised2018.11 | 41.7 | — | |
| SUM-GANsupEvaluation protocol=Canonical2018.12 | 41.7 | — | |
| SUM-GANAugmented Dataset=false2018.12 | 41.7 | — | |
| SUM-GANsupsplit=canonical2021.09 | 41.7 | — | |
| SUM-GANSupervision Setting=Supervised2021.12 | 41.7 | — | |
| Bi-LSTMEvaluation Protocol=Augmented2017.12 | 41.6 | — | |
| vsLSTMAugmented Dataset=true2018.12 | 41.6 | — | |
| SUM-FCNSupervision Setting=Unsupervised2021.12 | 41.5 | — | |
| Unsupervised Tessellationmode=unsupervised2016.12 | 41.4 | — | |
| DR-DSNLearning paradigm=Unsupervised2017.12 | 41.4 | — | |
| DR-DSNEvaluation Protocol=Canonical2017.12 | 41.4 | — | |
| DR-DSNLearning mode=Unsupervised2018.11 | 41.4 | — | |
| DR-DSNLearning Strategy=Unsupervised, LSTM hidden layer size=2562018.11 | 41.4 | — | |
| DR-DSNsplit=canonical2021.09 | 41.4 | — | |
| DR-DSNSupervision Setting=Unsupervised2021.12 | 41.4 | — | |
| SMILsplit=canonical2021.09 | 41.2 | — | |
| SUSiNetTask configuration=1-task, Augmented Dataset=false2018.12 | 41.1 | — | |
| Summary Transfer2016.12 | 40.9 | — | |
| Summary transferLearning Setting=Supervised2017.12 | 40.9 | — | |
| Summary transferLearning Setting=Supervised, Evaluation Protocol=Canonical2018.11 | 40.9 | — | |
| Summary Transfer2021.09 | 40.9 | — | |
| SUSiNetTask configuration=multi, Augmented Dataset=false2018.12 | 40.8 | — | |
| SEQ2SEQAugmented Dataset=false2018.12 | 40.8 | — | |
| Bi-LSTMEvaluation Protocol=Transfer2017.12 | 40.7 | — | |
| LSMO2021.09 | 40.3 | — | |
| Block Sparse2021.09 | 40.1 | — | |
| Long Short-Term Memory2016.12 | 39.8 | — | |
| Gygli et. al2016.12 | 39.7 | — | |
| SubmodularityLearning Setting=Supervised2017.12 | 39.7 | — | |
| SubmodularityLearning Setting=Supervised, Evaluation Protocol=Canonical2018.11 | 39.7 | — | |
| Gygli et al.Augmented Dataset=false2018.12 | 39.7 | — | |
| InterestingnessLearning Setting=Supervised2017.12 | 39.4 | — | |
| InterestingnessLearning Setting=Supervised, Evaluation Protocol=Canonical2018.11 | 39.4 | — | |
| CSUV2021.09 | 39.3 | — | |
| GANdppLearning paradigm=Unsupervised2017.12 | 39.1 | — | |
| GANdppEvaluation Protocol=Canonical2017.12 | 39.1 | — | |
| GANdppLearning mode=Unsupervised2018.11 | 39.1 | — | |
| GANdppLearning Strategy=Unsupervised2018.11 | 39.1 | — | |
| SUM-GANLearning Approach=Unsupervised2019.04 | 39.1 | — | |
| SUM-GANsplit=canonical2021.09 | 38.7 | — | |
| SUM-GANSupervision Setting=Unsupervised2021.12 | 38.7 | — | |
| DPP-LSTMEvaluation Protocol=Canonical2017.12 | 38.6 | — | |
| DPP-LSTMLearning Setting=Supervised2017.12 | 38.6 | — |